new evaluation of access to primary healthcare · 2010. 4. 14. · jeny shrestha february, 2010 ....
TRANSCRIPT
Evaluation of Access to Primary Healthcare
A Case Study of Yogyakarta, Indonesia
Jeny Shrestha
February, 2010
Evaluation of Access to Primary Healthcare
A Case Study of Yogyakarta, Indonesia
by
Jeny Shrestha
Thesis submitted to the International Institute for Geo-information Science and Earth Observation in
partial fulfilment of the requirements for the degree of Master of Science in Geo-information Science
and Earth Observation, Specialisation: (Urban Planning and Management)
Thesis Assessment Board
Chairman : Prof. Dr. Ir.M.F.A.M. van Meerseveen
External examiner : Prof. Dr. O. Verkoren
First Supervisor : Dr. S. Amer
Second Supervisor : Dr. J.A. Martinez
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION
ENSCHEDE, THE NETHERLANDS
Disclaimer
This document describes work undertaken as part of a programme of study at the International
Institute for Geo-information Science and Earth Observation. All views and opinions expressed
therein remain the sole responsibility of the author, and do not necessarily represent those of the
institute.
i
Abstract
Access to primary healthcare (PHC) has been a major goal of much health legislation and planning
processes in order to meet the healthcare need of population. Improvements in access to PHC pave the
way for advancement in the quality of people‟s life. However, efforts to conceptualize and
operationalize measures of access have varied. In this study, access is decomposed into five
dimensions as: availability, accessibility, affordability, acceptability and adequacy. One of the
motivations for this concept is the presence of relatively limited studies in access to healthcare
considering non-spatial factors like acceptability and adequacy. The relevance of access in this study
is substantive, in investigating for each of these dimensions in how far the distribution in space and
people is equitable.
Using primary data collected by household survey, variation in dimensions of access was measured,
by developing objective and subjective indicators, in three villages in the Province of Yogyakarta
(DIY). Variation of access was evaluated across different health facilities users and socioeconomic
classes using descriptive and explorative statistics. Existing situation of access to PHC was compared
with the related health policies in DIY. Using official census data of 2005, a comparative analysis
with primary data was done to complement the disaggregated data with aggregated one.
The results of this study showed that variation in access exists between villages and across different
socioeconomic class. Physical accessibility and affordability in general was not very problematic as
the result of effective policy implementation. In general people had no real preference over cultural
factors and gender of medical staffs (under acceptability). The analyses demonstrate that waiting time
in health facilities (under availability) and inter-personal treatment from medical staffs (under
adequacy) were important causes of dissatisfaction with access. These issues require policy attention
to further improve access to PHC. While developing health policies, consideration to internationally
accepted health standards can be effective in meeting minimum required standard for access to
healthcare. Results of analysis in scale effect revealed that aggregated census data tends to average
out variations as compared to data obtained at higher spatial resolution. Hence, care should be taken
before drawing conclusions.
In general, findings of this study indicate how dimensions in access can be quantified and measured
for the evaluation of access to PHC across different population groups. Also the possible effect of
scale in results of analyses is demonstrated. The approaches and findings of this study can be useful in
addressing problematic issues in access to PHC in the region.
Keywords: Primary healthcare, Access, Dimensions of access, Socio-economic stratification, Scale
effect
ii
Acknowledgements
This research is the result of valuable support from many institutions and individuals. Their
contributions in different ways have helped in the successful completion of this research.
I am very thankful to God who gives me strength and blessings every day of my life. My family
deserves special thanks for their love, support and faith in me besides the inspiration that they have
provided. I would like to thank Netherlands Fellowship Programme for the opportunity to pursue MSc
study in ITC which has broadened my academic knowledge and professional skills.
I would like to express my sincere gratitude to my supervisors, Dr. S. Amer and Dr. J.A. Martinez for
their guidance and critical suggestions from the inception till completion of this thesis. Their valuable
suggestions at various stages of this research were crucial for conducting my work in the right
direction. Their academic and moral support has been very inspiring throughout this study. I am also
thankful to all UPM staffs and teachers for the excellent academic guidance and assistance during the
programme. Special thanks to Ir. Mark Brussel for initial inspiration and introducing us to the city of
Yogayakarta. His guidance during fieldwork was very helpful.
I am very much thankful to „Pustral‟, Center for Transportation and Logistics Studies, Gadjah Mada
University for providing me necessary data and information, which were of immense value in this
study. Special thanks to my thesis advisor, Ir. Arif Wismadi for his suggestions and all administrative
support during fieldwork. I must thank Dr. Choirul Anwar, the Chief of Health Department,
Yogyakarta for his valuable time in providing me relevant information and documents. My
acknowledgement goes to Dimas, Pugo and Tamzil for their hard work in conducting field surveys
and also to all participants of the interviews. The hospitality shown by the city of Yogya and
especially all staffs of Pustral is highly appreciated.
My thanks go to all my colleagues and fellow Nepalese friends; Arun, Gopi, Diwakar, GRD, Ganesh,
Janak and Jay for their help, cooperation and good memories shared together during this course of
study. Special thanks to my friend Jiwan Limbu for encouragement throughout. All wonderful time
shared with friends and colleagues were valuable for providing an extremely diverse, interactive, fun-
filled and an intellectual rendezvous.
iii
Table of contents
1. Introduction ................................................................................................................................ 1
1.1. Background of Study .......................................................................................................... 1
1.2. Research Problem ............................................................................................................... 2
1.3. Research Objectives ........................................................................................................... 3
1.4. Research Questions ............................................................................................................. 3
1.5. Research Framework .......................................................................................................... 3
1.6. Structure of Report ............................................................................................................. 4
2. Review on Access to Primary Healthcare ................................................................................... 6
2.1 Introduction ........................................................................................................................ 6
2.2.1 Definitions and Concepts of Access to Primary Healthcare ......................................... 6
2.2.2 Dimensions of Access to Healthcare Service ............................................................... 7
2.3 Conceptual Framework for Access to Healthcare .......................................................... 11
2.3.1 Healthcare Utilization and Quality of Care ................................................................... 12
2.3.2 Equity in Access to Primary Heathcare ......................................................................... 13
2.4 Measuring Dimensions of Access to Primary Healthcare .................................................. 14
2.4.1 Developing Indicators to Quantify and Measure Dimensions of Access ..................... 14
2.4.2 Subjective and Objective Indicators .......................................................................... 15
2.4.3 Analyzing and Measuring Indicators ......................................................................... 16
2.5 Methodological Problems to be Resolved in Evaluating Access ....................................... 17
2.5.1 Effect of Scale in Analysis ........................................................................................ 18
2.6 Conclusion........................................................................................................................ 19
3. Study Area Description and Healthcare Policies ...................................................................... 21
3.1 General Description of Study Area ................................................................................... 21
3.1.1 Demographic Condition ................................................................................................ 21
3.1.2 Landuse and Economic Activities ................................................................................. 23
3.1.3 Administrative Units ..................................................................................................... 23
3.2 Health Policies and Planning Systems .............................................................................. 24
3.2.1 Decentralization ........................................................................................................... 24
3.2.2 Health Policies and Strategies ...................................................................................... 25
3.2.3 Health Organization System ......................................................................................... 27
3.2.4 Planning Process for Healthcare Service ....................................................................... 28
4. Research Methodology ............................................................................................................. 29
4.1 Research Design ............................................................................................................... 29
4.2 Fieldwork Preparation ...................................................................................................... 29
4.2.1 Study Area Selection ................................................................................................ 31
4.2.2 Sampling Strategy ..................................................................................................... 32
4.3 Field Work ....................................................................................................................... 32
4.3.1 Primary Data............................................................................................................. 32
4.3.2 Secondary Data ......................................................................................................... 35
4.4 Post Fieldwork ................................................................................................................. 35
4.5 Challenges During Fieldwork ........................................................................................... 35
4.6 Data Analysis ................................................................................................................... 36
iv
4.6.1 Socioeconomic Stratification of Sample Households ................................................. 36
4.6.2 Measuring Dimensions of Access to PHC .................................................................. 37
4.6.3 Synthesising Indicators .............................................................................................. 37
4.6.4 Access to PHC in Relation to Existing Health Policies .............................................. 38
4.6.5 Scale Effect in Analyzing Socioeconomic and Access Variation ............................... 38
5. Perceived Access to PHC at Micro and Macro Level ............................................................... 40
5.1 Household Characteristics and Socioeconomic Stratification ............................................ 40
5.2 Measuring Dimensions of Access to PHC .......................................................................... 45
5.2.1 Descriptive Statistics for Availability of PHC ........................................................... 45
5.2.2 Descriptive Statistics for Accessibility to PHC .......................................................... 47
5.2.3 Descriptive Statistics for Affordability of PHC .......................................................... 47
5.2.4 Descriptive Statistics for Acceptability and Adequacy of PHC .................................. 48
5.2.5 Overall Satisfaction Level with Access to PHC ......................................................... 50
5.2.6 Influencing Factors to Overall Satisfaction Level ...................................................... 51
5.3 Synthesis of Indicators to Develop Summary Scores .......................................................... 53
5.4 Existing Situation of Access to PHC from Policy Perspective ........................................... 58
5.5 Scale Effect in Analyzing Socioeconomic Attributes and Access to PHC.......................... 61
5.5.1 Comparing Individual Variables from Census and Primary data ................................ 61
5.5.2 Comparing Variation within Village with Variation at Sub district Level .................. 64
6. Discussions on Findings ............................................................................................................ 66
6.1 Sub-objective 1: Measuring Access to PHC at Micro Level .............................................. 66
6.2 Sub-objective 2: Existing State of Access in Relation to Health Policy ............................. 68
6.3 Sub-objective 3: Variations in Results Obtained from Census and Primary Data ............... 69
7. Conclusions and Recommendations .......................................................................................... 70
7.1 Conclusions ....................................................................................................................... 70
7.1.1 Main Findings from Sub-objective 1 .......................................................................... 70
7.1.2 Main Findings from Sub-objective 2 .......................................................................... 71
7.1.3 Main Findings from Sub-objective 3 .......................................................................... 71
7.2 Recommendations ............................................................................................................. 72
References ........................................................................................................................................ 73
Appendix A: Empirical references to indicators used in this study .................................................... 76
Appendix B: Description and rationales of indicators used in this study ........................................... 79
Appendix C: Content of household survey questionnaire .................................................................. 83
Appendix D: Content of Interview with Health Facility Personnel .................................................... 88
Appendix E: Distribution of summary scores of dimensions of access .............................................. 90
v
List of figures
Figure 1-1: Conceptual Framework .................................................................................................. 4
Figure 2-1: The health access framework [Source: Obrist, Iteba et al.(2007)] .............................. 11
Figure 2-2: Conceptual framework to evaluate access to PHC ...................................................... 12
Figure 3-1: Population growth trend in DIY (1971 – 2000) ........................................................... 22
Figure 3-2: Population density map of DIY .................................................................................... 22
Figure 3-3: Land use of DIY............................................................................................................ 24
Figure 3-4: Administrative boundary.............................................................................................. 24
Figure 3-5: Organizational Structure of Health System in Indonesia............................................. 27
Figure 4-1: Location of surveyed household in villages ................................................................. 33
Figure 4-2: Data collection and fieldwork observations ................................................................ 34
Figure 5-1: Proportion of socioeconomic cluster per village ......................................................... 43
Figure 5-2: General characteristics of socioeconomic clusters per village .................................... 44
Figure 5-3: Spatial distribution of socioeconomic classes ............................................................. 44
Figure 5-4: Percentage of subjective perception on waiting time in PHC ...................................... 46
Figure 5-5: Percentage of subjective perception on travel distance and travel time to PHC ......... 47
Figure 5-6: Percentage of subjective perception on total cost for PHC ......................................... 48
Figure 5-7: Percentage of subjective perception on factors related to acceptability and adequacy
........................................................................................................................................................ 50
Figure 5-8: Perceived satisfaction level with access to PHC .......................................................... 51
Figure 5-9: Synthesising indicators to develop summary scores for dimensions of access to PHC 54
Figure 5-10: Summary score chart for dimensions of access to PHC ............................................ 56
Figure 5-11: Standardized residuals for perceived satisfaction with various factors of access ..... 57
Figure 5-12: Geographic distance to health facilities per village in DIY, ...................................... 60
Figure 5-13: Population-doctor ratio per sub districts in DIY ....................................................... 60
Figure 5-14: Euclidean distance from households to heath facilities ............................................. 63
vi
List of tables
Table 2-1 : Empirical references to dimensions of access ............................................................... 10
Table 3-1: Population density per regency in DIY from years 2003 – 2006 .................................... 22
Table 3-2: General demographic characteristics of DIY ................................................................. 22
Table 3-3: National standards for health service, Indonesia .......................................................... 26
Table 4-1: Research Design............................................................................................................. 31
Table 4-2: Collected data from secondary sources ......................................................................... 36
Table 5-1: Statistics of different socioeconomic clusters................................................................. 42
Table 5-2: Descriptive statistics of waiting time in health facilities ................................................ 46
Table 5-3: Correlation matrices between perception on factors of access and overall satisfaction
level with access to PHC .................................................................................................................. 53
Table 5-4: Summary score for dimensions of access per socioeconomic class ................................ 58
Table 5-5: Common variables in census 2005 and household data ................................................ 62
Table 5-6: Distance to healthcare facilities from Census and primary data ................................... 64
Table 5-7: Socioeconomic and access to PHC variability in terms of CV ....................................... 65
vii
Acronyms
DIY Province of Yogyakarta
BPS „Badan Pusat Statistik’ Statistics of Indonesia
IDR Indonesian Rupiah
UGM Gadjah Mada University
PHC Primary healthcare
HSEC High socioeconomic class
MSEC Middle socioeconomic class
LSEC Lower socioeconomic class
GIS Geographic information system
WHO World Health Organization
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
1
1. Introduction
This study develops a methodology that contributes in the evaluation of access to primary
healthcare in the province of Yogyakarta (DIY), Indonesia. Study emphasizes on appraisal of
existing healthcare service, to see if variation exists in different dimensions of access. If the
variation is significantly high, then issues of disparity or inequality arise. This research aims to
highlight various dimensions influencing access, and in defining inequalities that arise from such
variation. Consistency in results obtained from different scale of analysis is also checked for the
study area.
1.1. Background of Study
Primary healthcare (PHC) service is an important concern for increasingly growing population
primarily in developing countries. Cities in the developing countries are experiencing unprecedented
growth for decades. The emergent population is aligned with the increasing demand for healthcare
service along with all sorts of other infrastructures and public service provision required to ensure the
basic quality of life. As the result, the situation might come when the existing infrastructure and
public facility will not be able to provide adequate service to the constantly expanding population. “If
population grows faster either through natural growth or movement in areas facilitated with more
services than others, it is possible that the per-capita measure of facility availability could be
worsened” (Yamauchi, Chowdhury et al. 2007, p.19). If continued, this process might gradually lead
towards significant variation further fuelling inequality in service provision.
Adequate level of access to PHC is a major health development issue. Improvements in PHC services
pave the way for advancement in the quality of people‟s life. Yamauchi, Chowdhury et al. (2007)
states that planners have potential role in providing better coordination in the service provision of
such facilities. Equal provision of basic public services like primary healthcare has been a matter of
interest for researchers, planers and policy makers since decades. “The achievement of equity in the
distribution of urban public facilities is a goal of paramount importance to urban planners, who must
analyze whether and to what degree their distribution is equitable” (Tsou, Hung et al. 2005, p.424).
Inequality in healthcare refers to disparities in access to health facility to large extend.
Access to PHC can be seen from a wide range of angles considering different factors and dimensions.
These factors depend largely on the objective of study and the context of study area. Number of
literatures (Penchansky and Thomas 1981; Millman 1993; Guagliardo 2004) have defined such
factors and dimensions as barriers to access while defining and conceptualizing access to healthcare,
which is explained in detail in next chapter. Factors like physical distance, travel time to reach the
facility, availability of transportation, service cost, waiting time, language etc are referred as barriers
because these factors impede people in certain socioeconomic condition to reach and achieve the
healthcare service.
Dimensions in access can be measured by developing indicators, both objective and subjective
indicators. Objective indicators are the observable facts and figures like distance, time, cost etc and
subjective indicators are normally derived from peoples‟ perception and satisfaction level on each
factors obtained from ground survey.
2
Scale of analysis should be considered carefully while analyzing variations in different dimensions of
access to PHC, as results of similar analysis may vary based on the selected areal unit. As spatially
aggregated data obtained from census are often used by researchers and policy makers to analyze
different contextual determinants of health, it is important to understand what factors and criteria
were used to aggregate such data. Aggregation of data might results in the loss of information which
might be relevant in specific study like healthcare. The limitations of scale problems are discussed
later in chapter two based on some empirical studies. This study analyzes and compares the results
from disaggregated primary data to the aggregated secondary data related to socioeconomic and
access to PHC in the study area.
In this study, influencing factors in access to PHC is measured in the Province of Yogyakarta,
Indonesia. Rapid urbanization in Indonesia is compounding the pressures for higher supply of
healthcare services, so is the case in the Province of Yogyakarta having four rural regencies and one
urban, the city of Yogyakarta. Regencies are further divided into sub-district and village being the
smallest administrative boundary. Depending on the physical and socioeconomic condition of people,
relevance of various factors is explored to evaluate the access to PHC. Results obtained from
aggregated and disaggregated data are compared to see the consistency in different scale of data.
1.2. Research Problem
As the result of rapid population growth and socioeconomic heterogeneity, variation in the provision
of PHC services can be assumed to occur. This affects the life of people in lower socioeconomic
group more adversely if the variation is significantly high. It is important to identify and address such
variations in order to provide adequate service to all people regardless of their socio-economic status.
Many studies have been done focusing on physical accessibility to define spatial equity in service
provision. However, there are other factors along with accessibility, which determines equity in
access to service. As mentioned earlier, there are different dimensions which influence the access to
primary healthcare service. However, it is difficult to decide which dimension poses more importance
across different socio-economic groups of people. For instance, geographic distance or cost of service
might not be a problem for affluent group of people who can afford their own vehicle. Quality of
service might be more relevant for them. On contrary, cost and long travel distance might be the
prime barrier for poor people to access such service.
This type of analysis can vary greatly depending on the spatial extent or scale of analysis chosen.
Variation in result of analysis gained from different scale of data is known as Modifiable Areal Unit
Problem (MAUP). According to Schuurman, Bell et al.(2007), researches on inequality context have
given significant attention toward the construction of socio-economic inequality indicators; however
less attention is paid in addressing the influence of scale. Stafford, Duke-Williams et al. (2008) stated
that the estimation of health inequalities in specific areas are determined by the way in which the area
boundary is defined. Effect of scale, in observing variation can be analyzed by comparing results of
similar analysis conducted for the same study area, which is divided into different smaller boundaries
separately. In Stafford, Duke-Williams et al. (2008) a metropolitan city is divided in three
neighbourhood boundaries separately for analysis. The area boundaries were selected based on
predefined census wards; physical features like rivers and main roads; and according to
socioeconomic homogeneity of residents. Although the result shows small difference (around 3%) in
this case, due to certain limitations like limited sample, authors clearly state the relevance of scale in
related studies.
Therefore, the research problem to be addressed in this study is to develop an appropriate method to
measure variations in access to primary healthcare considering different dimensions in access.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
3
Further, how to ensure if the level of variation in primary healthcare obtained from aggregated data
comprehends the result obtained from disaggregated one. Comparing the results obtained from large
scale aggregated data with disaggregated data shows the consistency and reliability of results from
different data sources.
1.3. Research Objectives
The main objective of this study is to evaluate access to primary healthcare, to observe if variation
exists in any dimension of access and if it meets the need of population.
The main objective is divided into three sub-objectives:
a. To develop an appropriate method to measure different dimensions of access to primary
healthcare to evaluate overall access.
b. To study how the planning for primary healthcare is done in context of Yogyakarta and to see
if existing situation matches with the planning standard.
c. To illustrate if information on socio-demographic and access to healthcare obtained from
aggregated data (census) correspond to the results obtained from disaggregated data (primary
household data).
1.4. Research Questions
For each sub-objective the questions are proposed to answer:
Questions for sub-objective 1
i. What are the appropriate methods to quantify and measure different dimensions of access?
ii. Do all dimensions have equal importance across different socioeconomic group of people?
iii. What is an appropriate areal scale for evaluating variation in access to PHC?
Questions for sub-objective 2
i. How is the planning for healthcare done based on health policies and national standards?
ii. Is the existing situation of access to healthcare in accordance to policy standards?
Questions for sub-objective 3
i. How different scale of analysis affects the result in mapping socioeconomic and service
variation in PHC?
ii. Does the result of analysis based on aggregated data, for larger spatial boundaries, matches
the actual situation within those areas?
1.5. Research Framework
The conceptual framework in Figure 1-1 shows the factors in macro level and dimensions in access,
considered to measure the existing situation of access to primary healthcare in DIY. The contextual
planning system and standards in health policies should be understood well prior evaluation of service
provision. Demographic information is relevant in this evaluation, as the need and perception on
primary healthcare service varies depending on the physical and socioeconomic condition of area as
well as population. Therefore various dimensions of access to primary healthcare are considered in
this study. By exploring these issues, dominant problematic factors causing variation in access to
4
PHC in context of study area can be found. This will highlight the important dimension in access to
be improved, in order to achieve better access to PHC to all.
1.6. Structure of Report
Chapter 1 - Introduction
This chapter gives an overall idea about the research structure starting with a brief background of
study. It comprises the research problem, main objective followed by several sub-objectives, research
questions to each sub-objectives and the conceptual framework of the study.
Chapter 2 – Literature Review
This chapter presents a brief review on definitions and concepts of access used in other empirical
studies which helped conceptualize access and its measures for this study. Detail definition of
different dimensions and conceptual framework in evaluating access to PHC adapted in this study is
explained. Variation in dimensions of access is reviewed in relation to the broader issue of equity.
Different methodologies used to quantify and measure access along with methodological problems has
been discussed.
Figure 1-1: Conceptual Framework
Problem Identification
- Dominant factors affecting access to primary healthcare
- Variation in different dimensions of access across different socioeconomic groups
- Planning and policy context
- Consistency in aggregated and disaggregated data
Suggestions and Recommendations
- Focusing dominant problematic factors under dimensions of access to further
improve existing situation of access to primary healthcare
Planning Context
Healthcare
planning
system
Planning
standards for
primary
healthcare
service
Demographic Data
Population
density
Heterogeneous
socio-economic
society
Access to Primary
Healthcare
Availability
Accessibility
Affordability
Acceptability
Adequacy
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
5
Chapter 3 – Description of Study Area
This chapter provides a general description of the physical, demographic and socioeconomic
condition in the Province of Yogyakarta (DIY). An overview on the effect of urbanization and
decentralization in Indonesia over provision of public services is discussed briefly. Healthcare
planning system, approaches in development of healthcare service and national policy standards for
access to healthcare in context of DIY is explained.
Chapter 4 – Methodology
This chapter describes the methodology to be carried out in this study. Detail information on research
design, data requirement, source of data and information and framework of analysis is provided. This
chapter is divided mainly into four parts; pre-fieldwork, fieldwork, post fieldwork and data analyses.
Preparation of questionnaires for household survey, base map, study area selection and sampling
strategy is explained in pre-fieldwork phase. Methods applied in carrying out ground survey,
interviews and data entry is reported in fieldwork phase. Data preparation, processing and analysis
formed the main part of post fieldwork data analyses. Various analytical methods in obtaining
answers to the research questions are described in this chapter.
Chapter 5 – Results
This chapter contains the results of analysis of variation in dimensions of access to primary
healthcare. This chapter highlighted the dominant factor or dimension affecting the overall perception
of individual on access to healthcare services. Findings of the comparative analysis between existing
situation and health policies in DIY are presented along with the results obtained consistency analyses
between aggregated and disaggregated data.
Chapter 6 – Discussion
The results of analyses, carried out to address all research questions, are discussed comprehensively in
this chapter. Critical discussion over strength and limitations of this study is also presented.
Chapter 7 – Conclusions and Recommendations
This chapter provides an overview on the concept, approaches and methods used in order to obtain
answers to all research questions of this study. Chapter concludes by highlighting main findings of
this research and further recommendations for future improvement of this study.
6
2. Review on Access to Primary Healthcare
This chapter comprises of review on definitions and concepts of access to different healthcares. It
gives detail explanation about different dimensions and conceptual framework in evaluat ing
access to primary healthcare services applied in this research. It also provides an overview on the
healthcare utilization as well as the quality of care, commonly used in evaluating health outcomes.
Issue of equity in access to healthcare has been discussed which help in clarifying the concept of
service variation and inequity in this research. This chapter further explains methodologies used
in measuring the dimensions of access by developing subjective and objective indicators. A brief
review on GIS based measures of access and methodological problems has been discussed. Issues
of different scale effects in analyses are discussed followed by conclusion of the entire chapter.
2.1 Introduction
Primary healthcare has been a major topic for many health studies, research and discussions of
medical care since decades. Increasing attention has been focused on this concern by medical
educators as well as state legislators in order to improve the supply of primary practitioners. However,
definition of the term „primary care‟ varies. Obtaining consensus on priorities or an agreement on the
content itself, has been a cumbersome task (Parker, Walsh et al. 1976). As stated in Bagheri,
Benwell et al. (2005) primary healthcare is an important step in providing „health for all‟, and is
widely acknowledged as a universal solution for improving population well-being by World Health
Organization and UNICEF. Guagliardo (2004) states that primary care is an essential form of
healthcare for maintaining population health as it is relatively inexpensive and easily delivered.
Author further explains that it is most effective in preventing disease progression on a large scale if
they are adequately provided in space.
Access to primary healthcare is considered as one of the indexes in achieving the goal of „health for
all‟ and it has different definitions depending upon different contexts. Different definition and
concept of access to primary healthcare is explained in the next section.
2.2.1 Definitions and Concepts of Access to Primary Healthcare
Access to primary healthcare can be seen from a broad perspective making it difficult to give a
precise definition. Efforts to conceptualize and measure access have varied depending on different
circumstances and context of study. Being a concern of social welfare, studies on access to healthcare
service have been carried out by different professionals like geographers, public health sector,
anthropologist etc. Thus the definition varies accordingly depending on the study approach. “The
most basic problem in defining „access‟ is that it is both a noun referring to potential for healthcare
use, and a verb referring to the act of using or receiving healthcare” (Guagliardo 2004, p.2). This
might create confusions due to overlapping of understanding between physical presence of primary
healthcare facility and ability and willingness of people in obtaining the care.
A more specific definition refers to ability to “secure a specified set of healthcare services with
certain level of quality, subjected to a specified maximum level of personal inconvenience and cost,
while in possession of a specified amount of information” (Oliver and Mossialos 2004, p.656). The
term „specified‟ in this definition, makes it easier for policy maker to define access depending on
specific circumstances for different places depending on the availability of resources to finance
healthcare. It can be summarized that general definition of access can guide the planner and policy
maker towards important factors, such as relevant range and quality of service, inconvenience,
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
7
disutility, cost, information dissemination etc, to be considered. If adopted and implemented properly,
this can serve as a standard against which existing access can be compared. Therefore this helps
policy maker to observe how improvement can be made or to check if they are improving access over
different area and population.
As various factors and issues should to be considered while defining access to primary healthcare,
some common factors can be grouped into different dimensions. Study of various dimensions and
factors affecting access to primary healthcare can further clarify the definition of access.
2.2.2 Dimensions of Access to Healthcare Service
Many authors have mentioned about different factors that impede access to primary healthcare. Due
to the complexity of access concept, it is important to look at each factor separately, even though they
are interrelated. Factors like availability of medical personnel, convenience to achieve health
services, actual use rates, service use in relation to some standards of need and consumer satisfaction
level with services has been highlighted by many studies, while exploring the overall access to
primary healthcare service. Dimensions of access differ with different geographical, socioeconomic
and cultural settings. Healthcare insurance, service cost, physical distance to reach the service, lack
of transportation, capacity of facility to serve the need of patients, indirect cost apart from health
insurance like travel cost, socio-cultural factors like race, language, gender etc and service quality
issues are some issues in access to healthcare service as documented in Millman (1993). Author has
grouped these issues into three dimensions; structural, financial and personal. Similarly, Penchansky
and Thomas (1981) has grouped those issues, termed as barriers, into five dimensions: 1. availability,
2. accessibility, 3. affordability, 4. accommodation and 5. acceptability; ‘5 A’. The first two
dimensions are spatial in nature. Availability refers to the total number of service from which user can
make their choice. Accessibility is related to travel impedance (time or distance) between spatial
location of user and services. The last three dimensions are non-spatial, related to cost, service quality
and cultural factors. Obrist, Iteba et al.(2007) also adopted above mentioned five dimensions in
clarifying the concept of access to healthcare but the term „accommodation‟ is replaced by
„adequacy‟ while explaining if patient‟s expectation towards quality of service and personal treatment
is met by the facility. The structural dimension in Millman (1993), possess the factors mentioned in
availability and accessibility, financial dimension covers the affordability and personal dimension
includes issues of adequacy and acceptability.
The concept of „5A‟ is used in this study while evaluating existing situation of access to primary
healthcare. Each dimension is explained and divided further into simpler quantifiable and measurable
form.
1. Availability refers to the extent to which a system provides facilities (which is the structural
form) and services (which refers to the process) that meets the needs of people (Campbell, Roland et
al. 2000). More than simple doctor-patient ratio, availability further deals with access to specific
gender of medical personnel, for example female general practitioner or nurse; access to medical
stores, laboratory or other equipments etc. Campbell, Roland et al. (2000) stated that organizational
access can be seen as sub-component of availability. This means even if people have adequate
physical access to the facility, there might be other factors creating barriers like length of time in
getting appointments, waiting time in before getting treatment or sometimes language barrier with the
facility professionals.
2. Accessibility is most commonly related with the geographic location of patient to the
location of facilities. Measures like spatial distance, travel time, mode of transportation used to reach
the facility, type to road network etc are considered assessing physical accessibility of people. Large
8
number of studies like (Talen and Anselin 1998; Black, Ebener et al. 2004; Bagheri, Benwell et al.
2005; Amer 2007; McGrail and Humphreys 2008) has been done on physical accessibility to public
facilities and mostly healthcare service. “Accessibility relates to the ability of people to overcome the
friction of distance to avail themselves of services at fixed points in space” (Amer 2007, p. 31). Study
of physical accessibility mostly incorporates three components: people; activities or service and mode
of transport to link them. The framework in Moseley (1979) states that accessibility varies according
to the characteristics of each of these components and it is influenced by the relationship between the
socioeconomic character of people, users, and spatial dimensions. More elaborative concept on
accessibility incorporates one more component, the temporal component along with the three
mentioned above. This deals with the moment of time at which the service is available or at which
people are able to participate. For example, the opening hour of the primary healthcare facility and
working hour of people in this study.
3. Affordability by its name itself refers to the financial component. There might be adequate
number of health facilities or medical personnel in an area, spatially close to the needy population.
But, these facilities might not be affordable to them. In such case, people tend to go to other facility
than the one closer to them, which provide subsidized rate or cheaper service. This dimension looks
into direct cost like doctor‟s fee as well as indirect costs like travel and medical costs that have affect
on overall access to healthcare. Other factors like possession and coverage of health insurance, public
supports such as subsidized rate provided for certain group of people like low income or elderly
population, are also incorporated in this dimension.
4. Acceptability deals with the cultural also religious factors of people. Factors like age,
gender, education level, race or ethnicity determines the level of acceptability of service provision to
large extend. For example, if the service available is socially acceptable by people like gender issue,
if people have some religious or cultural preferences towards choosing certain healthcare facility or if
the service provider and people use a common language to communicate. This also depends upon the
personal perception of people that might vary within a same religion or gender. Beliefs and
expectations of different groups of people should be considered while evaluating this dimension.
5. Adequacy is seen from two ways in this study: quality of service provided and personal
treatment by the service providers. Opinion about the medical treatment whether people trust the
medical ability provided by the facility or not, if they are satisfied with the quality of service or
personal behaviour of all facility personnel right from the point of entry to facility for example person
at reception till the end of medical treatment by doctors and laboratory personnel.
Acceptability and adequacy being subjective matter, it becomes complex to define a specific standard
or threshold line. These can be looked from the people‟s feeling, preference and perception on related
issues.
A list of empirical studies related to access and its dimensions, along with methodological approach,
that are reviewed for this study are summarized in Table 2-1.
These five dimensions can also be related to the concept of potential and realized access to healthcare
used in Andersen, McCutcheon et al. (1983) and Bagheri, Benwell et al. (2005). The potential access
looks at the ability of needy people in gaining healthcare in presence of healthcare facility in space. It
is important to understand the difference between “having access” and “gaining access” to healthcare
while evaluating the access to the service. Gulliford, Figueroa-Munoz et al. (2002) clarifies the
distinction between them, as the former may be the result of the availability of services and the latter
one refers to ability of individual to overcome financial, organisational and socio-cultural barriers to
utilise healthcare service.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
9
Some researchers tend to relate potential access with characteristics of the population (such as their
household income, education level, their attitudes toward medical care etc) and the characteristic of
the delivery system (like the type of organization, distribution of medical personnel and types of
facilities). Such characters aid in the understanding of meaning of access toward healthcare for
different socio-economic group of people. The realized access deals with actual utilization of
healthcare services. To simplify, potential access consider the factors, influencing the ability to enter
facilities before receiving the service and realized access deals with the perception of people towards
different aspect, such as quality of service, cost or personal treatment, after receiving the service.
In this study, availability, accessibility and affordability along with other socioeconomic
characteristics of people are looked from potential access to primary healthcare. Realized access
includes issues related to acceptability and adequacy. Once factors and dimensions of access are
determined, a general framework can be formulated to observe a stepwise sequence in evaluation of
access to primary healthcare.
Author and
year
Study Context and
Area
Dimensions in Access
Mea
sure
s
Operationalization /
Methodology
Av
ail
ab
ilit
y
Acc
essi
bil
ity
Aff
ord
ab
ilit
y
Acc
epta
bil
ity
Ad
equ
acy
/
Acc
om
mo
dati
on
Penchansky
and Thomas
(1981)
Concept of access in
health policy
United States
√
√
√
√
√
S / O
- Factor Analysis
- Multiple regression
- Correlation coefficient
Andersen et
al. (1983)
Access to medical
care
United States
√
√
√
√
-
S / O
- Descriptive statistics
- Correlation coefficient
- Multiple regression
- Factor Analysis
Fosu, G.B.
(1989)
Access to healthcare
in urban areas of
developing societies
Ghana
√
-
√
√
-
S / O
- Descriptive statistics
- Pearson correlation
- Multiple regression
- Standardized regression
coefficient
Guagliardo
(2004)
Spatial accessibility
studies in urban
areas
United States
√
√
-
-
-
O
-Two-step floating catchment
area
- Gravity model
- Kernel density method
(Black,
Ebener et al.
(2004)
Physical
accessibility to
health care
Honduras, Central
America
√
√
-
-
-
O
- Correlation analysis
- Regression analysis
Bagheri et
al. (2005)
Spatial accessibility
to primary
healthcare
New Zealand
√
√
-
-
-
O
- Drive time and least cost path
analysis model (network
analysis) in Arc Info 9.1
10
Author
and year
Study Context
and Area
Dimensions in Access
Mea
sure
s
Operationalization /
Methodology
Avail
ab
ilit
y
Acc
essi
bil
ity
Aff
ord
ab
ilit
y
Acc
epta
bil
ity
Ad
equ
acy
/
Acc
om
mod
ati
on
Omer (2006) Spatial equity
regarding physical
accessibility to
urban services (park)
Israel
√
√
-
-
-
O
- Proximity
- Container measure
- Minimum distance
- Moving average index
- Correlations income, religion
and obtained service area
Obrist, Iteba
et al. (2007)
Exploring and
improving access to
healthcare in
resource-poor
countries
√
√
√
√
√
S / O
- Outcome in terms of
health status
- Patient satisfaction
and equity survey
- Multivariate Analysis
Amer
(2007)
Spatial equity in
urban health services
planning
Tanzania
√
√
-
-
-
S / O
- Statistical analysis
- Pearson correlation coefficient
- Two step Cluster Analysis
- ANOVA
- GIS-based „what if‟
- Flowmap
Yamauchi,
Chowdhury
et al. (2007)
Spatial coordination
in public good
(education and
health facilities)
allocation Indonesia
√
√
-
-
-
S / O
- Descriptive statistics;
cumulative percentage,
distribution
- Non parametric regression
McGrail and
Humphreys
(2008)
Measuring spatial
accessibility to
primary care in rural
areas
Austrailia
√
√
-
-
-
O
- 2SFCA
- Network Analysis in ArcView
9.1, closest facility
Table 2-1 : Empirical references to dimensions of access
Summarized overview of the empirical literatures on evaluation of access and its dimensions to
public services (majority on healthcare service)
Note:
= Included in the literature
= Not included in the literature
S / O = both subjective and objective
O = Objective
Looking at this table it can be said that limited number of studies has considered non-spatial
dimensions like affordability, acceptability and adequacy while conceptualization and
operationalization of access to primary healthcare. Measuring spatial components like physical
accessibility and availability have been a common approach in evaluating access to healthcares.
- √
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
11
2.3 Conceptual Framework for Access to Healthcare
Different authors have developed their own models and framework in assessing access to healthcare
depending on the way they define it. Most of the literatures in healthcare give emphasis on the
outcome of service in the form of health status and equity on service. However access to healthcare
has been an important part in their framework. In the model of Millman (1993), dimensions of access
to healthcare is the first step which is followed by service utilization, quality and efficiency of service
and finally by the health outcomes. The framework adopted in explaining health access by Obrist,
Iteba et al.(2007) is shown in Figure 2-1. This framework shows the combined effect of macro level,
i.e. the policies and different healthcare services, and micro level, i.e. livelihood assets of people and
need of healthcare, on the dimensions of access. Obrist, Iteba et al.(2007) explained about livelihood
assets comprising of five popular capitals: human capital (like education level, skills or local
knowledge); social capital (social networks or affiliations); natural capital (natural resources like
land, water or livestock); physical capital (physical infrastructure like road and facility complex,
mode of transportation, facility equipments etc) and financial capital (income level, subsidies or
health insurance). Further it is explained that the availability of these assets is influenced by less
controllable factors like economy, political state, technological advancement or natural factors like
climate or hazards like flooding, draught or epidemics. These factors are referred as the vulnerability
state. Once access is gained, then the framework follows the similar pattern like Millman (1993).
This framework comprises both supply (health services) and demand side (health seeking behaviour)
and it places access in the broader context of livelihood assets. Interaction between; healthcare
services and organizational policy or institutions responsible to govern the services and; livelihood
assets that people can use during vulnerability contexts determines the extent to which access is
reached along the five dimensions.
With reference to these literatures, a conceptual framework to evaluate five dimensions of access for
this study was prepared, which is shown in Figure 2-2. Like in Obrist, Iteba et al.(2007), effects of
macro and micro level are observed on the dimensions of access to healthcare services. Factors in
Figure 2-1: The health access framework
[Source: Obrist, Iteba et al.(2007)]
Note: Red highlights the main focus of this literature
12
each dimension should be measured to evaluate access. Once access is gained and service utilization
is ensured, then people‟s perception on quality of care and satisfaction over each dimension can be
used to analyse most influencing dimensions over access and their relationship with the overall
satisfaction level of people.
Figure 2-2: Conceptual framework to evaluate access to PHC
Although this study is limited to evaluation of dimensions of access, an overview on the service
utilization and quality of care can be relevant to develop a clear concept on total healthcare system.
This will allow to compare the perception of individuals on quality and satisfaction over different
dimensions with those documented empirically in previous studies.
2.3.1 Healthcare Utilization and Quality of Care
Once need for primary healthcare is realized and access to service is acheived, then the quality of
service provided can be checked by evaluating the healthcare utilization rate. This is carried out in
Conceptual Framework to Evaluate Access to Primary Healthcare
A
C
C
E
S
S
Accessibility
Availability
Affordability Acceptability
Adequacy
5 A
Interrelation
Hierarchy of planning organization,
Health planning agencies, policies
and processes
Healthcare services
Hospitals, government healthcare
services, private clinics, traditional
healers, medical stores and others
M
a
c
r
o
L
e
v
e
l
Physical and Socio-economic
condition of people
(Livelihood Assets)
Need for Primary Healthcare M
i
c
r
o
L
e
v
e
l
Ensure access to primary healthcare
Findings
Patients‟ satisfaction with factors in
each dimension
Perceived variation in access
Feeling of equity in treatment
State of access at micro level in relation
to policies at macro level
Results
Influencing factors of dimensions in
access
Relation between satisfaction level with
each dimension and overall satisfaction
with access to PHC
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
13
the „Finding’ section in Figure 2-2. Obrist, Iteba et al. (2007) mentioned that, improved access and
healthcare utilization should be combined with good quality of care in order to get positive health
outcomes. The outcomes can be observed by measuring health status, patients‟ perception towards the
level of satisfaction and equity. Quality in healthcare is multidimensional. Quality can be evaluated
in respect with some predetermined standards (normative) or can be judged by more subjective way
like measuring patient‟s satisfaction level. For example, Maxwell (1992) have point out that results
obtained from the people opinion on satisfaction can differ from the result obtained by evaluating
technical efficiency of facility.
Campbell, Roland et al. (2000) defined quality of care from two principal dimensions; access and
effectiveness. Access, as already mentioned, confirms if people are getting adequate primary
healthcare at time of need and effectiveness confirms how effective is the care obtained. Quality of
care is defined as “the ability to access effective care on an efficient and equitable basis for the
optimization of health benefit/well-being for the whole population” (Campbell, Roland et al. 2000,
p.1617).
Quality of care has been discussed in more elaborative form in Maxwell (1992). Author has point out
some important components while defining the quality of care.
Effectiveness: ensures if the treatment provided is the best available in a technical sense and
evaluates the result of treatment.
Efficiency: compares if the healthcare output is maximised for a given input or vice versa.
For example comparing unit cost among different health facilities providing similar services.
Relevance: observes if the overall pattern of services is the best possible considering the
needs and expectations of population.
Equity: looks into issues like relative fairness in treatment. It ensures if some people or group
of people are being dealt less favourably.
These issues prove to be important while evaluating existing state of primary healthcare in my study.
Individuals‟ perception on issues like service quality, trust towards the ability of service providers and
feeling of equity has been addressed while evaluating their satisfaction level with existing primary
healthcare provision.
2.3.2 Equity in Access to Primary Heathcare
Achieving equity in access to primary healthcare has been a central objective to many health care
systems. Most governments have declared that citizens should get universal and equal access to good
quality of primary healthcare. The issue of equity should be addressed at macro level in the
framework shown in Figure 2-2, while developing policies for health care planning.
“The concept of equity refers to the degree to which services or amenities are distributed in an equal
way over different areas as well as economic, ethnic and political groups, with appropriate
consideration given to the needs of special groups such as children and the elderly” (Omer 2006,
p.254). Depending on context, policy makers give priority to different groups of people to secure
equal access to primary healthcare. For example such groups are commonly defined by income; social
status; geographical location; education level; ethnicity; gender; lifestyle etc. According to Andersen,
McCutcheon et al.(1983) equity of access is obtained when services are provided on the basis of
people‟s need. This refers to proportional equity based on needs. Authors state that inequity exists if
factors like race, income level or insurance coverage etc are important predictors of access.
14
Equity can also be seen from social aspect. According to Martinez (2005), to reveal the fact that
inequalities within cities really do matter, it is essential to consider an approach from social justice
perspective. Smith (1994) considers that justice involves treating people fairly, which in distributive
justice means that whatever is being distributed should go to people in the right quantities. Social
justice is concerned with the question of who gets what where and how, and more precisely who
should get what where and how.
Equity in this study is related with the variation in primary healthcare service received by people
across different socioeconomic classes. Further equity can be observed from peoples‟ perception on
quality of service and personal treatment they receive from healthcare facilities. Comparing similar
issues between different service providers, for instance government and private clinics, can further
clarify the equity concept in this study.
Once concept is clear and dimensions of access are determined, an appropriate method should be used
to measure each dimension in order to evaluate the state and (un) equal access to primary healthcare.
2.4 Measuring Dimensions of Access to Primary Healthcare
The need to measure each dimensions of access individually has been recognized in literatures in
order to evaluate the overall access to primary healthcare either with respect to certain norms or from
user‟s perspective. As the main objective of this study is to evaluate access to PHC from five
dimensions adopted from literatures like Penchansky and Thomas (1981) and Obrist, Iteba et
al.(2007), developing an appropriate method to measure each dimension becomes the main
methodological problem in this research. To measure each dimension of access, they can further be
divided into simpler quantifiable components. Developing quantifiable indicators of each component
is a commonly adopted method in measuring access to healthcare. To clarify the concept of indicator,
a brief definition, advantages and types of different indicators are explained in the following sections.
2.4.1 Developing Indicators to Quantify and Measure Dimensions of Access
Andersen, McCutcheon et al.(1983) have stated that there have been a number of summaries of the
research on the indicators that correlates with evaluation of healthcare service which should be
considered in various approaches while measuring access. Developing such indicators allows to
measure variation, if exists, in any dimension of access to primary healthcare across different socio-
economic groups. And if variation is found to be comparably high, then the indicators further enables
to measure inequalities in different aspect of access. As mentioned in Martinez (2005), advantage of
using indicators to measure inequalities is that they can communicate in a simple way while detecting
and quantifying such variations. Indicators can also be used in monitoring the area of priority for
policy intervention. The main functions of indicators can be summarized as to simplify complex
phenomenon, quantify them to measure and to communicate the results. Depending on functions and
purpose of policy, indicators are classified into three groups in Parnell and Poyser (2001) which is
also explained in Martinez (2005).
• Descriptive or baseline indicators: describes the existing situation of some system or process. They
are the initial data collected for each variable.
• Normative or target indicators: Once area of need is identified, target indicators help to set the
expected target or goal to be achieved. They allow evaluating and comparing the existing condition
with certain standard normally defined by policies.
• Performance or outcome indicators: enables to check if the targeted goals have been achieved from
policy side and also allows observing users satisfaction level with the obtained result.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
15
This study however is concentrated in developing descriptive indicators that exhibits the present
condition of access to PHC across different socioeconomic groups of people, considering different
dimensions of access. Development of indicators is focused on the needs, experiences and opinions of
people (users) towards the existing access to PHC. This helps in highlighting the problematic issues or
area of need to draw policy attention. Based on the concept of five dimensions of access, indicators
were developed referring to the questions raised towards each dimension in Obrist, Iteba et al. (2007).
To address the research question for sub objective 1, indicators were developed to distinguish
different socio economic groups of people to see variation in dimensions of access. Various household
characteristics can be summarized as indicators in order to come up with different socioeconomic
classes. List of socioeconomic and access indicators developed for this study are presented in
Appendix A with references from number of empirical studies in related field. These indicators can
be continuous data like income, crowding, employment dependency or age dependency and
categorical like education level, housing condition or ownership of assets. Detail description and
rationale behind all indicators developed for this study purpose are presented in Appendix B.
The evaluation of access can be seen from two ways: subjective and objective perspective. Therefore
subjective and objective indicators can be developed separately to measure them. The following
section will explain more about these indicators.
2.4.2 Subjective and Objective Indicators
Being a matter of social concern, policy makers in healthcare planning have been interested in both
objective and subjective approach in policy making process. Hence, developing subjective and
objective indicators becomes important in this approach. Large number of literatures like Veenhoven
(2002), Das (2008), Foo (2000) etc have stated the importance of both subjective and objective
indicators while evaluating the quality of life, where healthcare has been an important domain of
study. As explained in these literatures, objective approach focus on measuring „hard‟ facts like
income of family, education level, distance travelled to reach healthcare, cost paid for the service etc.
The subjective approach on the other hand measures „soft‟ matters like the people‟s opinion or
perception on some issue, their satisfaction level with income or the healthcare services. Subjective
indicators express individual‟s evaluation of objective matters. Subjective indicators are normally
measured using Likert scale. Likert scale is a commonly used scale in research surveys and
questionnaires related to subjective measures and respondents state their agreement to a level of
statement. For example, 5-point Likert scale for satisfaction can range from: very satisfied, satisfied,
neutral, dissatisfied and very dissatisfied. However there is no such defined range in Likert scale. For
instance Turksever and Atalik (2001) used a 4-point Likert scale in evaluating subjective quality of
urban life, while Das (2008), Foo (2000) and Ibrahim and Chung (2003) applied 5-point Likert scale.
Both of these approaches hold their own importance in evaluating social problems and developing
new policies. As mentioned in Veenhoven (2002), objective indicators provide information about the
actual state of social problems and the effects of attempts to solve these problems. Such information is
of indisputable nature or can be called objectively true which enables rational social action. Despite
of many controversies against subjective indicators, like being irrational that hampers scientific
management, Veenhoven (2002) stated that it is important in social policy making process for number
of reasons such as:
1. Need of social policy to consider peoples‟ psychological satisfaction and their support, as policy is
not limited to mere material matters. Therefore, subjective indicators are required to achieve these
subjective goals.
2. To evaluate the progress or outcome of policy intervention from subjective measurement.
3. Satisfaction level and preferences of people gained directly from people can better indicate
comprehensive quality of services provided.
4. The distinction between „needs‟ and „wants‟ of people for certain service can be measured with
subjective indicators.
16
In this study, objective indicators give information about the physical state of access to healthcare
like spatial distance to healthcare service, travel time, waiting time, direct and indirect cost to be
paid for service etc. Subjective indicator on other hand will provide information on the mental state
of people towards all these issues along with their satisfaction level with over all access to healthcare
service. Subjective indicators are concerned with individual‟s evaluations of various aspects of their
healthcare-seeking experience like the convenience to get service, cost, service provider behaviour,
or overall quality of the care they receive. For example, there might be a case when a person can
physically access a healthcare with ease but he might not be happy with the quality of service or
personal behaviour of facility personnel. There might be discrimination in medical as well as personal
treatment for different socio economic class of people which leads to dissatisfaction among certain
group of people. This kind of information related to personal feeling is gained from subjective
indicators.
It is important to observe link between results obtained from subjective and objective approaches, as
there are some contradictory conclusions about the relationship between these two approaches. For
example, Das (2008) exhibits a weak relation between subjective and objective approach in
measuring quality of life. However, Foo (2000) and Ibrahim and Chung (2003) recommend to use
both indicators in order to complement the limitation of specific indicators.
2.4.3 Analyzing and Measuring Indicators
The conceptualization and the measurement of both indicators have been a major concern in
healthcare research. According to Andersen, McCutcheon et al. (1983) multiple regression techniques
have been applied to analyze a range of potential and realized access indicators. Objective measures,
which normally refers to utilization rates, can be measured in different ways for example, simple
proportion of people visiting and not visiting a healthcare facility within a certain period of time or
total volume to service consumed. While analyzing subjective measures, some researchers have used
satisfaction indicators as a determinant of utilization and others relate it as a consequence of
utilization. Roghmann, Hengst et al. (1979) states that satisfaction measures are content oriented and
its validity is generally assumed or assessed through correlations with other variables like their
willingness to change the service providers if applicable. In such studies, regressions were computed
to predict satisfaction from utilization and vice versa. Other statistical methods like descriptive
statistics, multi regression, correlation and factor analysis can be applied to analyze the nature and
relation among different types of data collected on various aspects of study (Field 2005). Lahelma,
Martikainen et al. (2004) used logistic regression analysis to calculate inequality indices for health
service. Martinez (2005) stated that factor analysis and multivariate statistical techniques are the
most commonly used techniques in social research which are preferable approaches in measuring
socio-spatial variations. Twp step cluster analysis or K-mean methods enables to obtain different
clusters of socio economic classes using continuous as well as categorical variables as input data
These statistical analyses are relevant for this study for exploring and analyzing collected data on
both subjective and objective measures of access. These analyses will enable to obtain relation
between different dimensions of access and helps to validate the result obtained to evaluate overall
access to healthcare service.
Along with statistical techniques, application of Geographical Information Systems (GIS) has a great
importance in analyzing the observations and displaying results. Next section explains about the GIS
application in related type of research.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
17
GIS based Measures to Access
Increasing advancement in GIS in health organisations, together with the availability of data, has
supported studies related with developing measures of access to healthcare services. Black, Ebener et
al. (2004) stated that GIS is suitable in measuring spatial accessibility to healthcare as they enable
researchers to input, store, manage, manipulate of both spatial and attribute (textual) data, analyze
and visualise spatial information. Literatures like Bagheri, Benwell et al. (2005), Guagliardo (2004)
and many more have explained about measuring spatial dimensions of access; availability and
accessibility. Measuring straight line distance (Euclidean distance) and creating thiessen polygons are
some simple methods in assessing physical access. For more detail analysis, sophisticated network
analysis can be done calculating travel time, distance and mode of transport in GIS. Bagheri, Benwell
et al. (2005) calculated driving time to primary healthcare (destination point) from patients‟ house
(origin point) to apply least cost path analysis model using network analysis in Arc Info 9.1.
Higgs (2005) reviews number of literatures highlighting the use of GIS-based measures in exploring
the relationship between geographic access, utilization, quality of healthcare service and health
outcomes. These studies explore the spatial configuration of healthcare delivery system along with
service quality measures; role of transport system used to reach the care service for different socio
economic class of people and the characteristics of people or the area where they reside, seeking
healthcare in measuring access to healthcare service.
GIS software provides spatial analysts that offer excellent tools for spatial data management and
visualization. Amer (2007) has used GIS to identify and visualize trend like socially progressive or
regressive pattern in the distribution of healthcare facilities. GIS in combination with other software
like Flowmap is useful in identifying suitable location of facilities for certain number of facilities to a
defined population or territory. This can ensure the optimal distribution of service in space
considering the concentration of demand (people). Amer (2007) mentions that more sophisticated
GIS-based analytical techniques , for instance GIS-based „what if‟ in his study, can be used to
evaluate and support improvement of healthcare service in terms of spatial equity and efficiency
which can further support strategic planning of urban health services delivery.
As this study focus on spatial and non spatial components of access equally, a simple proximity
measures can be applied to examine the influence of spatial dimensions on total access to primary
healthcare service. Further observation on how peoples‟ perception of proximity to primary healthcare
facilities can be related to actual proximity to validate the conclusion of this research.
2.5 Methodological Problems to be Resolved in Evaluating Access
Main methodological problems mentioned in the academic literature related to equal access are
twofold. According to Oliver and Mossialos (2004) the first problem is to get consensus in the
development of specific definition of healthcare which will enable healthcare policy makers to make
policy that is more consistent in providing suitably (dis)proportionate access across different groups of
people with different level of needs. This refers to providing appropriate unequal access for unequal
need. The second problem is to develop an appropriate method to measure access. This requires the
formation of standard policy that specify explicitly issues like: 1) the minimum relevant range which
can be referred as a benchmark while evaluating the quality of healthcare services, 2) considerable
level of convenience for all healthcare users, 3) margin of cost to be paid in obtaining basic
healthcare and 4) the minimum amounts of information that people should have to get the advantage
of healthcare services.
However, getting total consensus in defining access to healthcare is a cumbersome job, as is it not
feasible to consider and fulfil the need of total population. Different individual might have different
18
perception and priorities about the concept and dimensions of access. More over developing universal
standard to evaluate access can be impractical as it is highly determined by the geographical location,
political condition to some extend and socioeconomic as well as cultural characteristics of users.
Homogeneity of area, defined by above mentioned factors should be considered while developing
social policies. Therefore, developing different standards for different homogeneous areas,
considering little tradeoffs to minority, might be appropriate for measuring and evaluating
dimensions of access in that particular area.
Another important methodological issue in related research in access is the effect of geographical
scale of analysis in the results. The affect of scale of analysis over results is discussed in the next
section.
2.5.1 Effect of Scale in Analysis
Spatially aggregated data often provided by statistics office are commonly used by researchers to
investigate the contextual determinants of health. Most studies use census areas as the geographical
units for convenience as detailed population data are available in this unit. Spatial extent or scale at
which the analysis is done has an effect on their values and results. According to Stafford, Duke-
Williams et al. (2008) results of analysis using aggregated data vary according to the selection of
areal unit which is well-known as Modifiable Areal Unit Problem (MAUP). Schuurman, Bell et
al.(2007) states that the complexity of scale in research analysis is of critical importance as large
number of deprivation indices on health outcomes are being developed for policy implementation.
Researchers use different indicators at different spatial scales to find area of service deprivation. This
allows to observe the relationship between socio-economic difference and variation in healthcare
service. This enables appropriate classification of high-risk populations, or the areas of deprivation to
inform policy makers. “MAUP refers to the problem that occurs when inferences, based on spatial
analysis, change when the same data are analyzed using either variations in administrative zoning or
through different scales” (Schuurman, Bell et al. 2007, p.596). The two main components of MAUP
mentioned in literatures are the scale effect and geographic boundary constraint.
Number of studies (Carstairs 1981; Cockings and Martin 2005; Haynes, Daras et al. 2007; Stafford,
Duke-Williams et al. 2008) has stated that automated zone design techniques can be an option to
counter the affect of MAUP. This helps to control the design of zoning system in order to create
robust aggregation of spatial information for the intended analysis to be undertaken. In automated
zone design, different zones are created by automated means in which zone boundaries are controlled
by statistical design rules. Zone design technique can be used to create zones by maximizing the
internal homogeneity of variable like physical or socio economic condition within each zone. A
potential application could be the testing of hypotheses of causal links between variables. For
example, if it is hypothesised that health condition in socially deprived area is worse than in well off
areas, one could aim to create zones which are homogeneous in terms of social deprivation. If the
hypothesis is true then strong resultant correlation could be expected between the independent
variable, deprivation in this case, and dependent variable, health condition, for the newly created
zones. Such zones might be created considering approximately equal population size at different
scales (Cockings and Martin 2005) or internal homogeneity in terms of environment or social
characters (Carstairs 1981; Haynes, Daras et al. 2007). Such zones are able to demonstrate stronger
relationship between variables than census units, and that larger areas produced stronger relationships.
Similar techniques were used in the 2001 England and Wales census to produce homogeneous output
areas (Martin, Nolan et al. 2001).
Haynes, Daras et al. (2007) grouped 1991 Census enumeration districts (EDs) using similar material
deprivation values , like housing type, in the zone design. Homogeneity of variables in each zone can
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
19
be measured by the intra-unit correlation (ICC) which can be calculated by dividing the variance
component between zones by the total variance. Stepwise multiple regressions can be used to identify
significant predictor variables.
From the review of these literatures, inappropriateness of spatially aggregated census data and ward
boundary for specific analysis like health condition or service can be assumed. It can be concluded
that automatically designed aggregation might reflect the actual situation for accurately than the pre-
existing administrative area boundaries. Rogerson (2006) explains about the inconsistency in results
of statistical analyses obtained from aggregated and disaggregated data. It is noted that “correlation
coefficients tend to increase with the level of geographic aggregation when census data are analyzed.
A smaller number of large geographic units tend to give a larger correlation coefficient than does an
analysis with a larger number of small geographic units” (Rogerson 2006, p. 165).
Effect of scale is relevant in this study, as it aims to check consistency in results obtained from census
and primary data. This will ensure the reliability of aggregated data over disaggregated one for
particular studies like socioeconomic variability and access to PHC.
2.6 Conclusion
This chapter reviewed the general definitions and concepts used to describe access to primary
healthcare service by various authors in different geographic context. As there is no such universally
accepted standard in defining and measuring access, commonly used dimensions of access was
explored which further developed into a model in measuring and evaluating overall access to
healthcare provision. This contributed in development of the conceptual framework for this study.
Although this study concentrates on the evaluation of different dimensions of access to healthcare,
review on different phases and aspects of healthcare outcome such as livelihood focus, healthcare
utilization, quality of healthcare and health related equity issues was very useful to develop an overall
idea on the primary healthcare system.
Access to healthcare can be measured by two approaches: subjective and objective approach. These
can be measured by developing subjective and objective indicators for each dimensions of access
respectively. In this research, indicators are the significant tools which simplify the dimensions in
such a form that they can be quantified and measured.
Overview on different statistical and GIS applications in related researches are relevant for
developing methodology for this study. Relation and dependency of variable in access can be
obtained from different statistical analysis. GIS can be effectively used: to explore spatial
characteristics of study area, in estimating proximity to healthcare facilities, to explore the spatial
variation in access to healthcare and finally to visual interpretation of the results.
The review showed the importance of access to healthcare, regardless to geographic or socioeconomic
context, in social policy making process. The output of this study is expected to highlight factors
influencing the access to primary healthcare and to display how spatial as well as social variations
exist in this regard.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
21
3. Study Area Description and Healthcare
Policies
This chapter presents a brief description on the study area and the planning process for
healthcare. The description includes the geographic location, surface area, administrative units,
and influence of decentralization, socio-demographic condition and other characteristics of the
Province of Yogyakarta. Overview on the procedure and indicators used by government planning
agency to develop policy standards related to health care service is also explained.
3.1 General Description of Study Area
The province of Yogyakarta is located in the South of Central Java Island. It is the second smallest
province out of 33 provinces, after Jakarta in Indonesia with the total area of 3185 km2. It is
surrounded by the province of Central Java and bounded by the Indian Ocean on south. A distinct
character of this province is that it still has its pre-colonial monarchy embedded in the administrative
structure. Yogyakarta is the only province that is led by kingdom Kraton Yogyakarta, besides the
formal government. Due to this reason it is known as the special region of Yogyakarta, Daerah
Istimewa Yogyakarta or DIY in Indonesian. Apart from monarchy there is democratically elected
legislative body, Regional People's Representatives Assembly known as Dewan Perwakilan Rakyat
Daerah. Javanese is the main ethnicity with 97% of total population. Majority of people in DIY are
city are Islamic with 91.8% followed by Christian (7.9%), Hindu (0.2%) and remaining Buddhist
(0.1%).
3.1.1 Demographic Condition
Based on the result of National Socio-Economic Survey of Indonesia in 2005, total population in the
province was recorded to be 3,281,800 in which 51% was female and 49% of male. Sleman regency
has the largest population and Kulon Progo is the least populous. The population density in 2005 was
1030 person per km2 with the growth rate of 1.88 percent. Between 1990 and 2000, Sleman regency
had a growth rate of 1.45% accounting for the highest rate in the entire province. This phenomenon is
supported by high population density in the municipality and compared to the others during that time.
Figure 3-1 presents the population growth trend in the province from year 1971 to 2000.
The highest population density at present is in Yogyakarta city that is around 13,000 people per km2
with area around 1% of total DIY area. In contrary, the density of Gunung Kidul regency is the lowest
with around 470 people per km2 with area coverage of 46% of the total DIY area. Figure 3-2 displays
the population density per village in DIY according to figures in census 2005 and Table 3-1shows the
population density of DIY per regency. The province has experienced significant growth in terms of
population in over 3 decades. Based on the result of National Socio-Economic Survey, the highest
percentage of DIY population by age is productive (16 – 60 years old) people. About 14% of the total
population are old, above 60 years. This shows that population of DIY tends to have a higher life
expectancy.
22
General exploration of demographic attributes of the Province of Yogyakarta is done in Arc GIS.
Census data obtained from the Statistics office of Indonesia (BPS) for the year 2005 was used as it is
the latest date for data availability. Information from census data is available on village level for the
whole province. Some general demographic information for five regencies in DIY are listed in Table
3-2.
Legend
Regency
Village
Population
Density
(person per
sq.km)
Figure 3-1: Population growth trend in
DIY (1971 – 2000)
Source: Statistics Indonesia, 2008
Figure 3-2: Population density map of DIY
Data source: Census, 2005
Regency Bantul Gunung Kidul Kulon Progo Sleman Yogyakarta
State Rural Rural Rural Rural Urban
Total population 800569 747782 447695 895408 515976
% of population in DIY 23.5 22 13 26 15.5
% of area coverage in DIY 16.5 46 18 18.5 1
Male (%) 48.5 48 49 49.5 51
Female (%) 51.5 52 51 50.5 49
No. of household 214558 178936 110867 236776 102716
No of districts 17 18 12 17 13
No. of villages 75 144 88 86 45
Table 3-2: General demographic characteristics of DIY Source: Statistics Indonesia, 2005
Regency Area (km2) Population Density
2003 2004 2005 2006
Kulon Progo 585 640 640 660 635
Bantul 505 1600 1610 1,625 NA
Gunung Kidul 1,485 460 460 470 460
Sleman 575 1635 1640 1660 1755
Yogyakarta 32 12,025 12,445 12,935 13,600
Total 3,185 1000 1010 1030 787.15a
Table 3-1: Population density per regency in DIY from years 2003 –
2006 Source: Statistics Indonesia Note: NA = Not available a = Not including Bantul regency
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
23
3.1.2 Landuse and Economic Activities
DIY has been divided into two major land types: rural and urban. All four regencies have been stated
as rural. Only city of Yogyakarta is defined as urban city by Indonesian Department of Public works.
Population density and economic activities are the main criteria in defining it as urban. Apart from
rural and urban land use, some areas are stated as preserved areas that include natural resources like
mountain range and water catchment areas. Figure 3-3 shows the main land uses in DIY.
The economic structure in DIY is boosted by public service sectors like educational institutions and
tourism. Agriculture, trade, and industry also have some contribution in economic development.
Tourism sector has the major contribution to the GDP (Gross Domestic Product); however,
agricultural sector employs the most people. Communication and transport sectors are gaining
prominence over years which aid to development to large extend. Main economic activities in urban
land use are trade and service while less populous rural areas are predominant by agriculture.
However steady conversion of agricultural land to build up form can be observed in the periphery of
Yogyakarta city, especially in Sleman on South. Shift from agriculture to other economic sectors has
lead to tremendous growth in such sectors. This growth is attributed to huge amount of migrant
population from areas outside the province. High rate of rural urban migration has resulted in high
population density with widespread unemployment and lower living standard. Unemployment is
problematic but with the rate fluctuating between 5–8%, which is comparatively lower than in other
regions in Indonesia. These phenomenon leads to a heterogeneous society, dominantly in the city and
its surrounding areas.
3.1.3 Administrative Units
There is a descending level of administrative subunits in the government administration process.
There are twenty seven provincial level units which are divided into districts (kabupaten), sub-
districts kecamatan) and further into villages (desa or kelurahan). Village is the lowest tier of the
administrative hierarchy. The nation is centrally governed from Jakarta since independence, from
where national level decisions like line of authority, budget allocation, personnel appointments are
done. Regional and local governments enjoy little autonomy. Their role is mostly administrative like
implementing policies and regulations. They essentially serve as subordinate administrative units in
local level, which support the functional activities of Jakarta based departments.
The DIY has four districts also known as regencies; Sleman, Bantul, Gunung Kidul and Kulon Progo
and one city, the city of Yogyakarta which is the capital of the province. City of Yogyakarta is
located centrally, surrounded by Bantul and Sleman on the South. Gunung Kidul is the largest
regency with surface area of 506.80 km2 where as the City of Yogyakarta has the smallest surface
area of 32.5 km2. The administrative units of DIY are shown in Figure 3-4. Rural regency has a
regent and mayor in case of city of Yogyakarta, who holds the authority for the local development of
public service sectors like education and healthcare facility. Next administrative unit is the sub
district which also has a head person who has certain level of power in development and decision
making processes. The sub district heads are directly accountable to the mayor or regents. Similarly
there are village heads and village head office in each village, where village development plans or
other service related problems are discussed before proposing to the higher administrative levels.
24
3.2 Health Policies and Planning Systems
Overview on the decentralization in Indonesia is useful for this study as the planning process and
policies for healthcare service follows the regulations after decentralization. Significant influence of
decentralization can be seen on public service sector including healthcare.
3.2.1 Decentralization
Decentralization is the devolution of political decision making to local governments and
communities. The Indonesian Ministry of foreign affairs has defined decentralisation as “a means to
hand over political, financial and administrative authority from central to local (regency/city)
governments, so that the government can facilitate and guarantee better public services for the
people.” Indonesia experienced decentralisation in 1999 with the passage of certain laws on Regional
government (Law number 22/1999) and on fiscal balance between the centre and the regions
(Number 25/1999). This restructured the administrative framework by providing local government
the responsibility for planning and provision of public services. Central government is however
responsible to monitor and evaluation such development processes. Decentralization is viewed as a
positive development towards stronger association and transparency between people and public
service developments. One advantage is that authority at local level is more aware with the ground
situation and has greater knowledge about the service related problems. So they can meet the needs of
people more efficiently, also being easily accountable by local people. By the distribution of power,
better representation of local people is assumed to be achieved ensuring maximum public
participation in development processes. It is further believed to enhance effective distribution of
public services. Empirical studies on public services in Indonesia, documented in Yamauchi,
Chowdhury and Dewina (2007), have demonstrated that physical accessibility to public services like
school and hospitals has improved during decentralization period. But, per capita availability of
school and local medical healthcare, puskesmas, have decreased over time. The authors mentioned
that despite of the coordination in spatial allocation of such services, availability is coming up as a
problem due to mobility of endogenous people and higher growth rate. These phenomena partially
cancel the advantages of the coordinated efforts on public service allocation. There are some
drawbacks of decentralization in public service delivery. Yamauchi, Chowdhury and Dewina (2007)
states that investments in public service can be biased depending on local income and endowment
which complicates the coordination of investment decisions across different socioeconomic
communities. However it falls outside the scope of this research.
Figure 3-4: Administrative
boundary Data source: Census, 2005
Legend
Figure 3-3: Land use of DIY Data source: Census, 2005
Legen
d
Built forms
Regency
Forest
Agriculture
Green land
Water body
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
25
3.2.2 Health Policies and Strategies
The government of Indonesia has been redeveloping policies to improve the health service throughout
the nation. One of the policy visions in the line is the creation of “Healthy Indonesia 2010”. As
health is considered as a shared responsibility, the policy forces the Ministry of Health and Social
welfare to develop collaborative relationships with other parties involving all strata of community,
other related government agencies and private sector. As stated in WHO (2009) the “Healthy
Indonesia 2010” is developed with the main goals:
1. To initiate and lead a health orientation of the national development
2. To maintain and enhance individual, family and public health along with improving the
environment
3. To maintain and enhance quality, accessible and affordable health services
4. To promote public self-reliance in achieving government health
This study is related more to third and fourth goal. In the mean time, with the introduction of two new
Acts on Local Governance (Act No. 22/1999) and on Financial Balance (Act. No 25/1999), the
implementation of decentralization policy in Indonesia was emphasized from 2001. This empowered
provinces and regencies with autonomy in formulating policies and decision making in health service
considering local needs. It is believed that the service can be provided more effectively and urgent
problems of the area can be resolved faster. Collaboration between the National Health Development
and decentralization policy agreed on four paramount strategies serving as the pillars in formulating a
Strategy for National Health Development, which are:
a) Concept development of health
b) Professionalism
c) Community health maintenance assurance (Jamkesmas) and
d) Decentralization
The third strategy is relevant for the scope of this study as it deals with the „affordability‟ dimension
in evaluating access to primary healthcare. It focuses on community health insurance programs like
Jamkesmas. The aim was to guarantee equity in access to health services and the service quality. This
program was based on the concept of adequacy, equity, efficiency and effectiveness.
Indonesian government has been trying to increase the healthcare for the poor people. According to
the document provided by the health agency in Yogyakarta, in year of 2009 Department of Public
Health raises the number of poor people who receive subsidy for health insurance premium from 36
million people (2005) to 60 million people (2009). Health insurance premium was also increased
from 2.1 trillion (Indonesian Rupiah) to 3.7 trillion. By this increment, 60 million poor people are
expected to receive free healthcare by bringing the insurance card to government hospital and local
government clinic, puskesmas. As this policy is explicitly formulated for poor people, it is important
to understand how „poor‟ is identified. For social stratification, numbers of socioeconomic indicators
are developed by the Planning Office at regency / municipality level in context of local area and
living standards. The indicators includes attributes like income, employment status, physical
condition of house, nutrition, possession of consumer goods like telephone, television, motor bike, car
etc. According to the planning officer in Yogyakarta, the criteria for deciding different
socioeconomic class of people does not focus on mere income, as considering the possible affect of
26
inflation in future. It is assumed that with certain level of income, one might not be able to maintain a
similar kind of living standard in future, the way they are living at present, if the market price rises.
These indicators are developed in reference to the census data. The central Bureau of Statistic (BPS)
is responsible to provide data needed for similar planning procedure and for implementation. BPS
data is used also for the financial allocations to districts and in estimating „quotas‟ for poor people.
The social data produced by BPS is collected through surveys like National Social economic survey
(SUSENAS). Such survey is designed in order to collect social population data which is relatively in
wider scope. Collected data includes information on education, health/ nutrition, housing /
environment, criminality, social culture actions, consumption, income of households, possession of
vehicle type and other consumer goods and household welfare level. A team is set at sub-district level
for the survey in identifying poor households. When a household is considered to be poor after
evaluating with those indicators, then it will receive a letter known as „Surat Keterangan Tidak
Mampu‟. With this letter of poor the household receives free or highly subsidized rate at public
services. With the possession of this letter, one becomes eligible to get a health card known as „Katu
Sehat‟. The local authorities provide the lists of qualifying individuals and districts are allocated
quotas for health cards, based on the estimated percentage of poor people residing in that district.
Such cards cover the service cost of both outpatient primary care in puskesmas and free treatment at
hospitals (generally at third class public hospitals).
A nominal rate of Rp. 5,000 (about US dollar 0.55) per month per card holder is set as the standard
contribution rate. At present there are different types of free health insurance premium like Jamkesda
(district level health insurance for poor), Jamkesmas (civil health insurance for the poor), Askeskin
(Indonesia health insurance for the poor) etc depending upon their area and employment status. The
main function of these insurance is the same as previous health card that is to provide subsidized
health service and free in case of primary healthcare.
The national standards for healthcare service related to this study are presented in Table 3-3.
Service Area Indicators Service Standards Quality
Scope Services
Health care
facilities
-Distribution of
healthcare facilities /
coverage of health
services
For neighbourhood
unit population
should be less than
30,000
-At least 1 unit of
puskesmas / sub
district
-1 unit of health post
(posyandu) / village
-1 unit medical post
/ 3000 people
-1 unit of child and
maternal healthcare
(maternity hospital /
BKIA) / 10,000 to
30,000 people
-1 unit puskesmas /
120,000 people
-1 unit hospital /
240,000 people
Location of
facility should be
in centre of the
sub-districts /
districts
Should be located
in clean area far
from disease
sources , garbage
dumping sites and
pollutions
Table 3-3: National standards for health service, Indonesia
Source: Standard guidelines of minimum service level, 2001
The health policies seem to give priority to the spatial distribution and service cost. The report
(Moeloek 1999) from the Ministry of Health documented that in spite of adequate number and
distribution of heath facilities, the health services are still below standards in terms of service quality
as it is far from peoples‟ expectation. Another problem mentioned is regarding the uneven distribution
of health service personnel, inadequate educational quality and unbalanced health manpower
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
27
composition which results in longer waiting time by patients in healthcare centres and low service
performance. Although, there is not any officially documented policy regarding this issue,
information related to waiting time in healthcare facilities was obtained from the interview with head
of health planning agency and puskesmas personnel during fieldwork. According to them, the
indented optimal waiting time is considered to be around 30 minutes. However, these issues need
further policy attention.
3.2.3 Health Organization System
After implementation of decentralization, 349 regencies and 91 municipalities (also referred as
district/regency/city) became the key administrative units. One level down in administrative
hierarchy is the sub district, which has at least one public health centre, puskesmas headed by a
medical doctor. It is usually supported by two or three sub-puskesmas, generally headed by nurses.
The medical doctors occasionally visits the sub-puskesmas for instance once in a month for
monitoring. Most of the puskesmas are equipped with four wheel ambulances for emergency cases
also to support mobile health service to provide service to underserved populations in urban as well as
remote rural areas. At village, the smallest administrative unit, there should be at least one integrated
family health post. These health posts are established by community people with the help of medical
assistant from health centres. The aim is to provide preventive healthcare service also to promote
healthcare awareness. Midwives, both professional and traditional, are deployed to the villages
commonly for maternal and child healthcare.
The hierarchy of organization of health system starts with the Ministry of Health at the national level.
It is then followed by health planning agencies known as dinas kesehatan at province and further at
district levels. Then puskesmas are established at sub district level which is followed by sub
puskesmas, midwives and health posts at the lowest administrative unit (village). The flowchart of
organizational structure of health system is shown in Figure 3-5.
MOH level from central to peripheral level
Ministry of Health, Central level
Departemen Kesehatan (DEPKES)
Provincial health office, Province level
Dinas Kesehatan Propinsi (Dinkes)
District level health office, District level
Dinas Kesehatan Tingkat Kabupaten / Kota
Sub district level health center
Pusat Kesehatan Masyarakat (Puskesmas) tingkat Kecamatan
Village level health center
Posyandu
Sub Health Center Village Midwife Clinics Integrated Health Post
Figure 3-5: Organizational Structure of Health System in
Indonesia
28
3.2.4 Planning Process for Healthcare Service
Decentralization has provided a platform to support more community participation and involvement
of different stakeholders in the development and planning process for public services. Both local
communities and government organizations are often involved in the process of decision making for
public goods investments and service allocation. This has encouraged the bottom up planning system.
When people realize a need of certain healthcare service, villagers coordinated by the village office
with a village head approach the health office at sub district level and then to district level with a
proposal of investment project for that service. Then the district health office set a team of survey for
the evaluation of validity of the need of that project. In presence of many project proposals, the
district office selects a good project based on priority analyzed from the survey and thorough
analysis. Then it allocates fund to the selected project, received from central government allocated
for that region. However this process has some problems like difficulty in coordination and lengthy
decision making period. As both local communities and government are involved in the decision
making process for different projects, gaining consensus is a very difficult and lengthy process. And
even with the government‟s effort to coordinate investments across communities in their jurisdiction,
it cannot be guaranteed that the investments will be well coordinated to gain equal access of the
service by all people. Hence attention is given in the coordination for an effective and faster project
approval, planning and implementation to meet the demand of people.
Conclusion
This chapter provided general information on demographic condition in DIY such as population
density, population growth trend, urban rural land use types and hierarchy of administrative
boundaries. Such information was relevant for getting familiarize with the study area and also
important in selection of areas for primary data collection. Brief explanation on health policies and
planning process helped in understanding the healthcare system at macro level. This was important as
one of the objectives in this study was to compare existing situation of access at micro level with the
policy standards in the area context.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
29
4. Research Methodology
This chapter contains the methodological approach to address the research questions of this study.
Answers to the research questions are obtained by following a method that has five stages. In the
first stage, literature and empirical studies of access to healthcare provided a firm base to develop
initial concept of methodology. Research design is developed to determine the required data,
source of data and methods of obtaining them. The second stage is pre-field work preparation for
the fieldwork. General study about the study area along with the overview of available secondary
data and information was carried out. In relation with the research question and possible
availability of secondary data, questionnaire was prepared for household survey and interview
with primary healthcare facilities. The field work of 28 days was carried out during which all
relevant primary and secondary data was collected in the third stage. All collected data is
processed and checked for its consistency in the fourth stage which is the post field work stage.
The final stage is the data analysis. Methods for analysing both subjective and objective attributes
are applied to study the existing state and variation in access to primary healthcare in the study
area.
4.1 Research Design
The research design of this study is shown in Table 4-1. It presents details on data required, source of
data collection, methods to be applied to obtain answers for research questions to achieve each sub-
objective. The research design also includes the type of analysis carried out in this study
4.2 Fieldwork Preparation
A base map of study area was prepared with the available spatial data on administrative boundary and
geo-referenced Google image. Various indicators were developed to quantify and measure five
dimensions of access identified after literature review. Indicators developed under each dimension of
access are presented in Appendix B along with description and rationale. List of all relevant
information and household questionnaire for primary data collection was prepared accordingly. The
questionnaire consisted of two major sections: first was the general information related to household
and their socio-economic characteristics, second part focused on factors related to the access to
healthcare facilities. This section included type of healthcare facilities respondent visited and their
perception on satisfaction with various attributes of access to the facility. The questionnaire
comprised questions on both subjective and objective attributes of access. From objective attributes
answers for travel time, travel distance, cost, income level etc were obtained. Subjective attributes
included personal perception of level of satisfaction with various dimension of access to health care
facilities.
The questionnaire was developed in a simple structure which can be understood easily by
respondents. Questions related to subjective attributes and ranking options were translated in the local
language, Bahasa in order to be filled by the respondent themselves. An interview format was
prepared to obtain information from health facilities personnel.
30
Sub-
objectives
Research
Questions
Required data Data source Methods Analysis
1 - What are the
appropriate
methods to
quantify and
measure different
dimensions of
access?
-Do all
dimensions have
equal importance
across different
socioeconomic
group of people?
- What is an
appropriate areal
scale for
evaluating
variation in access
to PHC?
Literatures,
Empirical studies
Census data
Socio-
demographic
characteristics
Spatial factors
(geographic
location and
distances)
Individuals‟
perception
BPS –
Statistics of
DIY
Household
survey
Center for
Transportation
and Logistic
Studies,
Gadjah Mada
University
(Pustral,
UGM)
1. Developing
indicators for
socioeconomic
characteristics of
respondent
2. Developing
indicators for each
dimension of
access
3. Measuring
perceived level of
satisfaction with
each dimension
and overall access
4. Comparing
perception on
satisfaction level
between different
economic classes
of respondent.
5. Synthesising
indicators to
developing
summary scores
for dimensions
Descriptive
statistics; mean,
range, standard
deviation
percentage
count
Two step cluster
analysis
GIS; Proximity
measure
Chi square test
Correlation
coefficient
Coefficient
matrix
Plotting
summary scores
in multi-
dimensional
charts
Standardized
residuals
3
2
-How is the
planning for
healthcare done
based on health
policies and
national
standards?
-Is the existing
situation of access
to healthcare in
accordance to
policy standards?
Health policy
and planning
standards for
healthcare
facility in DIY
Demography data
Number of
existing
healthcare
facilities
Spatial factors
(geographic
Planning for
Health
Department
(Dinas
Kesehatan)
Regional
Development
Agency DIY
Province
(Kepatihan
Yogyakarta)
BPS –
Statistics of
Studying planning
process for
healthcare service
in DIY
Comparing
predefined
standards with
existing situation.
Ratio of
population to
healthcare
facilities and
Total number of
doctors
Distance
proximity
measure
Evaluating
individuals‟
perception over
policy measures
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
31
location and
distances)
DIY and their
implementation
4
3
-How different
scale of analysis
affects the result
in mapping
socioeconomic
and service
variation in
PHC?
-Does the result
of analysis based
on aggregated
data, for larger
spatial
boundaries,
matches the
actual situation
within those
areas?
Census data:
Aggregated
socio-
demography
data;
information on
access to
healthcare;
spatial data
Disaggregated
data on similar
issues from
primary source
BPS –
Statistics of
DIY
Pustral, UGM
Household
survey
Comparing
aggregated census
data with
disaggregated
household data.
Evaluating
consistency in
statistical analysis
Descriptive
statistics;
percentage
count
Coefficient of
variation
Correlation
coefficient
Table 4-1: Research Design
4.2.1 Study Area Selection
For the selection of study area, units of administrative boundaries were considered. The province of
Yogyakarta consisted of five regencies (districts), 77 sub districts and 438 villages. For this study,
village was an appropriate unit of observation since it was the basic unit of consensus-building for
bottom up approach in decision makings. The scale of analysis was chosen also in concern to the
secondary data to be obtained, as data were available on such administrative boundary. Certain
criteria were developed for selecting sample villages for detail study. Population density,
socioeconomic heterogeneity and spatial location of villages were considered for selection.
Percentage of prosperous household figure presented in census and visual interpretation using Google
image was used to determine heterogeneity in socio economic condition. All regencies in DIY
followed similar health policies formulated at the province level, regardless to different socio-
demographic structure in rural and urban regency. For instance, according to one of the health
policies, there should be at least one puskesmas in each sub district. However it did not mentioned
about the geographic area or population size of such sub districts. As area and population of sub
districts in rural and urban regency has vast difference, a comparative analysis of access to PHC
between rural and urban area was intended in this study. City of Yogyakarta being the only urban
regency, it was selected purposively. Two villages Tegalpanggung and Kricak were selected from
city of Yogyakarta. Tegalpanggung had the highest population density in the whole province and
located in the centre of city of Yogyakarta, hence was selected. Kricak was located at extreme
northern boundary of the city with comparatively higher population density than other villages
located in boundaries. Reasons to consider geographic location was to observe variation in access to
PHC, if any, in highly crowded city core and growing city fringe. Another spatial characteristic of
these villages was the river and major road network defining the physical boundaries. Dense informal
settlement was found along the river bank in west boundary of both villages. In contrast to this, more
formal basically commercial activities were found along main road lines. Such characteristics were
assumed to provide the possibility of exploring heterogeneous group within each village. Villages
32
were selected also considering the geographic area coverage and percentage of population covered by
those villages in relation to the respective sub districts they belonged to. Selected villages had the
largest area and population size among other villages within that sub district, hence was considered to
be representative for that sub district. This criterion was checked as the third sub objective of this
study was related to analysis of scale effect. Hence for comparative analysis between aggregated data
for sub districts and disaggregated data at village level, most representative village were chosen.
Sleman was selected as rural regency as it had highest population density among four rural regencies
in DIY. Village Tridadi was chosen from Sleman considering higher population density and
heterogeneous mixture of formal and informal settlement observed from Google image.
Administrative support provided to carry out survey and interview in related health organizations in
this village also played some role in the selection. After selection of villages, sampling strategy was
designed to carry out household survey. The main objective in this strategy was to collect data which
was representative of the population in village as far as possible.
4.2.2 Sampling Strategy
Depending on time constraint and limited financial resource, total sample size for this study was
determined to be around 300 households from three selected villages. This was about 3 percent of
total number of households residing in all three villages. Therefore proportional number of household
from each village was calculated considering the same percentage. Random sampling was applied for
the selection of households within those villages.
After the purposive selection of villages and determination of sample size, random sampling was
applied for the selection of household for survey. Digitization in ArcGIS was done to define the
cluster of built forms within each village. The digitized pockets excluded the major road and vacant
lands as far as possible. The built form in two villages in city of Yogyakarta was more dense and
clustered together. Therefore, grid of 100 meters by 100 meters was laid over the digitized pockets
and then calculated numbers of random points were created over grid boxes in ArcGIS. These points
were pre determined for each village, i.e. 3 percent of total household in the village. Digitization and
random points were also laid for the third village. However, grid was not applied due to the sparsely
located smaller pockets of built forms. Number of sample points within each digitized pocket was
decided according to the area covered by those pockets. Location and random points of sample units
for household questionnaire is shown in Figure 4-1.
4.3 Field Work
The required data for the study was collected from two main data sources; primary and secondary
sources during fieldwork.
4.3.1 Primary Data
Household survey was carried out in the selected villages along with three surveyors in their local
language. Objective data and individual perception about different aspect of the primary healthcare
facilities was collected through structured questionnaire from 28th September 2009 to 9th October
2009. Prior to the real household survey, the questionnaire was well explained to the surveyors and
required translation of certain portion in questionnaire was done. A pilot survey was carried out in
one study village to confirm the surveyors understanding of the questionnaire, also to check the time
length required to complete one questionnaire. The criterion for household survey was that the
respondent must be the head of family or the spouse of the head of family. This criterion was set so
that the respondent should be able to give the detail information about the whole household members,
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
33
for example overall income level. It was assumed that the head of family or spouse can reflect the
perception or feeling of all other household members.
100m X 100m grid Tegalpanggung
Kricak
Tridadi
Figure 4-1: Location of surveyed household in
villages Note: Image obtained from Google Earth-Pro, acquired in 2007
34
Ten questionnaires were filled in pilot survey. The time required to complete one questionnaire was
about 20 to 30 minutes. Certain changes were done after the pilot survey in order to further clarify the
questions. Surveyors were given training in reading map over laid with satellite image, in order to
find the spotted house on image for survey. In case of unavailability or unwillingness of respondent in
the spotted house or if the spot is on land use type other than residential, the survey was carried out in
the nearest house from the spot. Second part of the primary data collection was interview with
personnel at two health planning agency, Dinas Kesehatan at City of Yogyakarta and Sleman regency
and with medical staffs at 5 public healthcares, Puskesmas, in three sub districts containing the
sample villages. For this purpose, an official request letter was issued from the University of Gadjah
Mada and a permission letter from the municipality was obtained. Then mentioned health planning
agencies and puskesmas were visited along with a local translator in the appointed date. Purpose
behind the interviews at health planning agencies was to gain understanding about the health planning
process and related policy structure in study area. And purpose of interviews at puskesmas was to
understand type of health services being provided and to query about common problematic factors, if
any, in access to healthcare reported by patients. Some photographs from fieldwork are displayed in
Figure 4-2 .
Figure 4-2: Data collection and fieldwork observations
(a) Household survey with the help of local surveyors
(b) Socioeconomic heterogeneity with in village (Kricak)
Well of households Poor informal settlement along river bank
(c) Data collection from health planning agency and public healthcare facility, puskesmas
Health planning agency Waiting area in Puskesmas
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
35
4.3.2 Secondary Data
Census data for the whole province of Yogyakarta for year 2005 was obtained from the Center for
Transportation and Logistics Studies, Gadjah Mada University. This data was provided by the
Statistics office of the province. Spatial data like administrative boundary of study area for the whole
province along with road network data was also obtained from the same organization. Census
contained number of socio-demographic information at village level for the whole province.
Knowledge on the planning system for health facilities was gained from the visit to regency planning
office and health agency. Few supporting documents on numbers of different healthcare facilities and
medical doctors were also acquired. Table 4-2 displays the list of secondary data collected during
field visit.
4.4 Post Fieldwork
Data collected from primary source, questionnaire and interviews, is processed, checked for
consistency and entered in digital form. After entering, every 10th household data was cross checked
with the filled questionnaire, to ensure the accuracy in entering. As most of the collected documents
were in local language, Bahasa, required translation was done. As the census data contains a huge
number of attributes related to socio demographic information, infrastructures, agriculture and other
public services, only limited attributes relevant to this study were selected and translated in English
for further use.
4.5 Challenges During Fieldwork
Some difficulties and constraints were faced before and during the data collection.
Lack of sufficient data and information on socioeconomic condition and health facilities
hindered the selection of villages and finalization of questionnaire before field visit.
Due to absence of high resolution satellite image, preparation of base map for sampling for
household survey took time. Processing high resolution image from Google pro and geo-
referencing and preparing mosaic from small pieces of images was time consuming.
Spatial location of healthcare facilities was not obtained as per expectation. Location of five
puskesmas, in three sub districts containing selected villages, and hospitals being visited by
respondents were spotted in Google image with the help of secondary document collected
from health planning agencies and with the help of local spatial knowledge. Personal visit
was made to the spotted health facilities for conformation.
Language barrier was another difficulty during fieldwork. As a consequence, survey had to be
done with the help of local surveyors and translators.
Documents obtained from planning and health agency was in their local language.
Translation required extra time and effort.
Sample size had to be limited due to time and financial resource constraint.
Due to the Muslim festival „Idul Fitri‟ fieldwork had to be stopped for a week. During the
festival period all related organizations were closed for interviews. Also due to unavailability
of local surveyors even household survey could not be carried out.
36
Type of Data Description Data condition Source
Spatial data Administrative boundary;
Province, regency, sub-district
and village
GIS data (vector) Pustral, UGM
Road network GIS data (vector) Pustral, UGM
River GIS data (vector) Pustral, UGM
Demography data Population data, geographic
socioeconomic data
Excel
Document (hard copy)
Pustral, UGM
BPS- Statistics of DIY
Health Facilities
Number of health facilities per
village (public hospitals,
puskesmas)
Excel
Document (hard copy)
Regional development
agency DIY
(Kepatihan Yogyakarta)
Number of medical doctors per
village
Excel
Document (hard copy)
Health Agency
(Dinas Kesehatan)
Puskesmas
Standards in health policy and
information on regional health
insurance
Document (hard copy) Health Agency
(Dinas Kesehatan)
Table 4-2: Collected data from secondary sources
4.6 Data Analysis
4.6.1 Socioeconomic Stratification of Sample Households
Analysis of socioeconomic strata of respondent was relevant, as the major part of this study was based
on the subjective perception of respondent on dimensions of access to PHC, which was assumed to be
influenced by their socioeconomic characteristics. Also socioeconomic attributes such as possession
of letter of poor and government health card, explained in section 3.2.2, were important as being
directly related to the dimension affordability. Number of socioeconomic attributes collected from
household survey was combined together in order to categorize each sample households into different
socioeconomic groups. Description and rationale for each socioeconomic indicators used is presented
in (Appendix B).
In this study different attributes of household characteristics were given equal consideration in
evaluating their socioeconomic status rather than limiting to a single variable like income. Amer
(2007) explained that using single variable like income or education to distinguish social strata
oversimplifies the result and cannot reflect the actual reality. The author also highlights the negative
possibility of using scoring method in which variables are assigned certain weights based on some
assumption, which are later combined to come up with one value as the final result. This approach has
some drawbacks, as the process of assigning weights is preconceived and subjective approach.
Therefore to avoid such possibilities, Amer (2007) used a multivariate and exploratory stratification
approach, Two-Step Cluster analysis. This analysis is adapted in this study for similar purpose.
Two-Step Cluster analysis is an exploratory tool designed to reveal natural groupings within a data
set. The algorithm used in this process offers number of beneficial features like 1. ability to create
clusters using both categorical and continuous variables, which is suitable for the data set in this
study; 2. it enables automatic selection of the number of clusters based on similarities and
dissimilarities within and between clusters. It also allows user to specify the number of clusters if
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
37
needed and 3. it can efficiently analyze large data set. Clusters can be characterized based on the
descriptive statistics provided for each input variable. After classification, GIS is applied to visualize
the spatial distribution of households from different socioeconomic classes within each sample
villages, to see socioeconomic heterogeneity at village level.
4.6.2 Measuring Dimensions of Access to PHC
Descriptive statistics was used to measure both objective and subjective indicators developed under
dimensions of access to PHC. The first three dimensions of access to PHC; availability, accessibility
and affordability, was measured from objective as well as subjective approach by developing
indicators. The remaining two dimensions; acceptability and adequacy, being subjective in nature
only subjective indicators were developed to measure perception of respondent on related issues.
During household survey, questions were asked about objective and/or subjective aspect on each
dimension of access to PHC. The subjective perception over each indicator and overall satisfaction
level with access to PHC was measured using 5 point Likert scale that ranged from very satisfied to
very unsatisfied. Statistics like mean, standard deviation and minimum – maximum range was used to
compare variation in objective indicators such as travel distance, time and waiting time in health
facilities between different villages, health facility users and socioeconomic classes. For travel
distance, Euclidean distance was measured in GIS between sample household points and visited
public health facilities as reported by respondents. Descriptive statistics, i.e. cumulative percentage of
respondents is calculated to compare the subjective satisfaction level with each indicator and over all
access to PHC.
For categorical data, cross tabulation table was used as it is a basic technique for examining the
relationship between nominal or ordinal variables. It also offers a test of independence and measures
of association between such variables using Cramer‟s V statistics. If the frequency of subjective
opinion in Likert was less than 5 in cross tabulation, then 5 point Likert scale was aggregated into 3.
This was done by considering „satisfied‟ and „very satisfied‟ as one value indicating positive opinion
and „dissatisfied‟ and „very dissatisfied‟ was replaced with one value indicating negative opinion.
Neutral opinion was left unchanged.
Correlation matrix was computed between levels of satisfaction with various indicators and overall
access to see the significance and strength of correlations. Factors which had highly significant strong
correlation can be considered as better predictors of overall satisfaction level with access to PHC.
4.6.3 Synthesising Indicators
Summary scores were developed for all dimensions of access by synthesis of underlying indicators, as
it is hard to deliver a meaningful message from indicators alone when looked from a broader
perspective of policy making. Developing a summary score by synthesising indicators by considering
their relative importance into a single value has been a common practice which enables easier
interpretation for policy maker. “It is the analysis of indicators against the wider context and policy
objectives that provides the added value of converting information into intelligence” (Wong 2006,
p.81). This will provide a more explicit evaluation of the state of access to PHC following the
concept developed for access in this study.
Considering equal importance of all dimensions in access, equal weights were applied to each
indicator while developing a summary score. Also as there was no reference of subject matter expert‟s
opinion on prioritization of different indicators, equal weights were applied. The concept followed
while developing indicators based on their relevance referring to literature Obrist, Iteba et al. (2007)
can fairly justify the use of equal weights for this study.
38
Box plot was used to see the nature and range of summary score distribution. Following Wong (2006)
average value of summary scores for each village, health facilities and socioeconomic classes were
plotted in multi dimensional radar chart for visual interpretation. As the result of equal weighting,
information on dominant problematic factors might get lost while developing summary scores. Hence
referring to Amer (2007), standardized residual was used to see variation in occurrence of satisfaction
level with each indicators under dimensions. Residuals give the difference between observed
frequency of cases with satisfaction and the expected frequency, specifying the deviation from the
expected average number. It helps to observe how proportion of respondents from different cases
(villages or using different health facilities) differed based on their satisfaction with each indicator.
4.6.4 Access to PHC in Relation to Existing Health Policies
One of the objectives of this study was to compare the existing situation of healthcare service in terms
of access, with health policies and planning standards in study area. This enabled to compare the
existing situation of access at micro level from people‟s perception with planning norms at macro
level. As discussed earlier in section 3.2, after decentralization, provinces and regencies were
empowered with autonomy in formulating policies for public facilities including health. Hence health
policies and standards formulated at provincial level were referred for descriptive analysis.
Information on spatial distance to health facilities and number of doctors working for public health
facilities, obtained from census data was used to prepare choropleth maps to display spatial variation
in geographic distances from villages to nearest public health facilities in DIY. Also maps were
prepared to display relative variation in population doctor ratio referring to the Indonesia‟s average
population doctor ratio obtained from Dash (2000).
Descriptive statistics, i.e. cumulative percentage, was used to compare the existing state of factors
related to access to healthcare, which are also mentioned in health policies of DIY.
4.6.5 Scale Effect in Analyzing Socioeconomic and Access Variation
To achieve answers to the research questions related to the effect of scale in analysis, a comparative
study was done between primary and secondary data source. The purpose was to explore the
consistency of the results obtained from aggregated and disaggregated data. Referring to different
techniques to minimize scale effect (refer section 2.5.1) from empirical studies (Carstairs 1981;
Hyera 2003; Stafford, Duke-Williams et al. 2008), this study was concentrated on socioeconomic
homogeneity of three selected villages to observe variation in socioeconomic characteristics and
access to healthcare within each village. Common variables related to socioeconomic and access to
PHC was selected between primary data and census for comparison.
Descriptive statistics like percentage count, mean, minimum-maximum range of the common
variables was used to see how well census figures at village level represented the actual condition
within villages obtained from primary data collection.
Variation observed at village level was compared with the variation at sub district level using similar
variables from census data. Referring to Turksever and Atalik (2001), variability in socioeconomic
attributes and access to PHC was explored in terms of coefficient of variation (CV) at sub district and
village level. Coefficient of variation was computed by dividing the standard deviation of each
attributes by its mean and it indicated absolute variation within sample villages. This will further aid
in analysing variation obtained from different sources of data. In all applicable cases, CV was
computed using aggregated village data from census to observe variation at sub districts level and
disaggregated household data for variation within each sample village.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
39
Referring to empirical studies on scale effect, it was assumed that correlation coefficients tend to
increase with higher level of geographic aggregation. “A smaller number of large geographic units
tend to give a larger correlation coefficient than does an analysis with a larger number of small
geographic units” (Rogerson 2006, p.165). Hence to ensure consistency of aggregated and
disaggregated data in statistical analysis, correlation coefficient was checked between similar
variables of access from different source of data.
40
5. Perceived Access to PHC at Micro and
Macro Level
This chapter presents the results of analysis in four main sections. First section includes the
general characteristics of the sampled households, emphasizing socioeconomic attributes.
Measurements and analysis of indicators on each dimension of access to PHC is presented in the
second section. The results of the analysis of perceived satisfaction level with each dimension
including variability in their multiple indices are presented in this section. Third section evaluates
the existing state of access comparing with health policies and standards in DIY. The last section
contains the comparative analysis of common variables between census and household data to
show the consistency of using aggregated data in this type of study.
5.1 Household Characteristics and Socioeconomic Stratification
Household characteristic
General characteristics of 273 sampled households were explored to get an idea of the collected data
prior to running other analyses. Majority of respondents were female and about 75% of households
were composed of 3 to 6 members. As the survey was carried out in day time between 9 am to 6 pm
and most of the households have male as working members. As the province of Yogyakarta is
renowned for educational institutes, literacy rate was found higher than the national literacy rate
which was 71% in 2008 according to Statistics of Indonesia (BPS), also larger than the ratio for the
province itself (65% of literacy). Person above 15 years old who have attained formal education and
are able to read and write is defined as being literate by BPS. About 79 % of sample households were
found to attain education from senior high school up to university level education. The household
monthly income was classified into 5 ranges starting from 700,000 Indonesian Rupiah (IDR).
Absolute income poverty line for province of Yogyakarta was defined as 194,830 IDR per capita for
year 2008 by BPS, which is equivalent to around 21 US dollar per month. This was less than 1 dollar
per day per capita if referred to Millennium Development Goal poverty indicator. Using such
standard poverty line could not reflect the actual state of reality, as it differs from country to country.
Hence, poverty thresholds or indicators should be chosen as per the context of study area. Range of
income was designed assuming average household size of four members considering the income
poverty line by BPS. About 38% of households stated to be in the range below 700,000 IDR out of
which 65% of households have more than 4 members. The percentage of households below income
poverty line in the province was 18% in 2008 as stated by BPS. Around 67% of households reported
to have the letter of poor from the local government. Majority of the interviewed houses (52%) were
semi permanent with half non brick or wall and non plastered floor. This shows a significant
difference with the figure stated by BPS for 2008, which was 13% of semi permanent houses in the
province. Definition of each category of construction type was referred from the census questionnaire
from the BPS (presented in Appendix B). Regarding basic infrastructures like water source and
sanitation, majority of households with 58% used well water and around 77% had private toilets with
septic tanks. These figures are similar with the percentage presented by BPS for the whole province
of Yogyakarta.
From these figures, it can be said that the study area have larger proportion of poor households as
compared with the statistics for the province provided by BPS. However, it is in better state in terms
of literacy rate when compared with the provincial as well as national literacy rate.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
41
Socioeconomic stratification
A number of socioeconomic indicators were used as input variables in the Two Step Cluster Analysis
to assign each household a socioeconomic class. Prior to performing cluster analysis, independence of
variables was explored by checking relation between them to ensure if there is any strong association
between variables. Number of iterations was done entering and removing different socioeconomic
variable to come to a reasonable result of classification. Finally, cluster analysis was performed using
thirteen independent variables. As the analysis produced only two automatically formed clusters, a
specific number of clusters was entered to get three classes of socioeconomic strata for this study.
Three clusters were identified that distinguish well off and poor households with middle
socioeconomic class in between. The result is summarized in Table 5-1.
The first cluster was composed of 43% of total sampled households. This cluster had the lowest
percentage of university level education attainment and all households with no formal education fell
in this cluster. There was a distinct difference in the proportion of monthly income range and
possession of letter of poor in this cluster. About 89% of households in this cluster had monthly
income in the lowest range of 700,000 IDR with about 80% of households possessing the letter of
poor. This cluster had the highest percent of households which does not have house ownership and
with temporary housing structure among three clusters. The ratio of bedrooms to total household
members was the lowest however the difference is insignificant with cluster 3. Average value of age
dependency (ratio of member from age16 - 60 to total household members) was lowest among three
clusters. Similar was the case with the mean value of ratio of employed members to total household
members. The mean value for the number of assets like bicycle, motorbike, telephone and refrigerator
owned by households in this cluster was also lowest although the difference is less between the third
cluster. From this statistics, cluster one was considered to belong to lower socioeconomic cluster
(LSEC) among three.
Cluster two was composed of 35.5% of total sample households with largest percentage (42.3%)
having university education attainment. Monthly income range was distinctively higher and
comparatively lower portion of households with the letter of poor. The result of house ownership
status was contradicting from other indicators, as cluster three had the highest percentage of
households with house ownership. However, significantly large portions of houses were with
permanent structure with better physical condition and larger bedroom ratio. It can be assumed that
living condition can be better in such house than in poor structured with ownership. It was also
supported by the result of highest portions of houses with private toilet and none of houses without
toilet. The average values for the ratio of working age members and employment ratio to the total
household members were higher than other two clusters. Also the mean values of all household assets
were large with all households having car in this cluster. This cluster presents the better living
condition and higher socioeconomic characteristics, therefore considered as higher socioeconomic
class (HSEC).
Most of the statistical results of the last cluster were in between cluster one and two which was
composed of 21.6% of total sample households. This cluster had similarities in results with other two
clusters in case of different indicators. For instance, none of the household without education and
with lowest range of income was in this cluster, which is similar to cluster two. Also the combined
percentage of higher education attainment is close. On other hand, the value for permanent structure
of houses, type of toilet, working age ratio, employment ratio, bedroom ratio and asset possessions
were similar to that of cluster one. Therefore, this cluster was considered to be in better
socioeconomic condition than the first cluster and worse than the second, so it was considered as
middle socioeconomic class (MSEC).
42
Indicators
Overall
frequency %
Cluster characteristics (Frequency %)
1(LSEC)
N=117 (42.9%)
2(HSEC)
N = 97 (35.5%)
3(MSEC)
N = 59 (21.6%)
Cate
goric
al
varia
ble
s
Education level
No education
Below senior high school
Senior high school
University
Others
0.5
19.5
41
31
8
1
35
42
22
0
0
5
34
42
19
0
12
43
37
8
Range of monthly income
Less than 700,000
700,000 – 1,400,000
1,400,000 – 2,800,000
2,800,000 – 5,600,000
Above 5,600,000
38
39
19
4
0
89
10
1
0
0
0
42
47
11
0
0
88
12
0
0
Possession of letter of poor
Yes
No
68
32
80
20
54
46
64
36
House status
Owned
Not owned
78
22
61
39
85
15
100
0
Type of house construction
Permanent
Semi permanent
Temporary
39
53
8
13
69
18
86
14
0
14
81
5
Toilet type
Public
Private
No toilet
18
77
5
27
66
7
2
98
0
27
65
8
Indicators Overall mean Cluster mean
(Higher value indicates better socioeconomic condition)
Con
tin
uou
s varia
ble
s
Household characteristics
Age dependency (16 - 60
years / total members)
Employment ratio
0.59
0.37
0.54
0.28
0.66
0.52
0.57
0.31
Housing Condition
Bedroom ratio
0.54
0.47
0.66
0.49
Asset possession
Cluster mean and % with asset possession
(Higher value indicates better socioeconomic condition)
Number of bicycles
Number of motorbikes
Number of cars
Number of telephone
Number of television
Number of refrigerator
1.01
1.24
0.07
1.38
1.15
0.55
0.79 (57%)
0.68 (56%)
0 (0%)
0.94 (50%)
0.91 (80%)
0.21 (21%)
1.32 (67%)
2.08 (93%)
0.2 (19%)
2.1 (90%)
1.52 (99%)
0.99 (76%)
0.95 (65%)
0.97 (73%)
0 (0%)
0.95 (56%)
1 (93%)
0.47 (48%)
In this way, each household was classified into three different socioeconomic strata which were used
for the subsequent analyses that follow. The proportion of each socioeconomic class for the sample
Table 5-1: Statistics of different socioeconomic clusters
[Note: LSEC = lower socioeconomic class; MSEC = middle socioeconomic class; HSEC= higher socioeconomic class;
N = number of households]
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
43
villages is shown in Figure 5-1. The portion of LSEC household was found to be greater (almost
48%) in village Tegalpanggung. This village had the highest population density in DIY and is located
in the core of Yogyakarta city. Employment opportunity could be one reason for the poor household
to concentrate in the city core, which avoids the travelling cost for work. Also the west boundary of
Tegalpanggung is defined by Chode River, along which informal settlement resides. Kricak had the
highest portion (40%) of HSEC households. However, difference in proportion between HSEC and
LSEC was not much (only 3.5%). Tridadi also had highest portion of LSEC households (44%)
followed by HSEC households (37%). Kricak and Tridadi are steadily growing villages in terms of
population and economic activities, as located in the periphery of Yogyakarta city. As observed
during field survey, the majority of physical living condition (housing condition) was found better in
Kricak and Tridadi with less temporary houses with bigger living area. During survey it was noted
that better off households preferred to live in urban periphery that offers bigger space, as the core
being highly dense. Also they normally owned a private vehicle, generally motor bike, so
transportation was not a big problem to reside at considerable distance from city core.
Figure 5-2 shows some of the general characteristics of different socioeconomic clusters and Figure
5-3 displays the spatial distribution of different socioeconomic classes in three villages.
Middle socioeconomic class
Lower socioeconomic class
Higher socioeconomic class
43.8
%
47.8
%
36.5
%
22.8
%
18.8
%
29.4
%
37.4
%
23.5
%
40
Figure 5-1: Proportion of socioeconomic cluster per village
(b) Education level (a) Income range
HSEC
MSEC
LSEC
HSEC
MSEC
LSEC
HSEC
MSEC
LSEC
44
Figure 5-2: General characteristics of socioeconomic clusters per village
Figure 5-3: Spatial distribution of socioeconomic classes
Tegalpanggung
Kricak
Tridadi
HSEC
MSEC
LSEC
HSEC
MSEC
LSEC
HSEC
MSEC
LSEC
(c) Housing structure
HSEC
MSEC
LSEC
HSEC
MSEC
LSEC
HSEC
MSEC
LSEC
(d) Toilet type
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
45
5.2 Measuring Dimensions of Access to PHC
Depending on the nature of indicators developed, each dimension of access to PHC was measured
and compared using descriptive statistics. This allowed to observe variation in state of access from
different dimensions between households from different villages, visiting public and private health
facilities and belonging to different socioeconomic classes.
5.2.1 Descriptive Statistics for Availability of PHC
For availability, questions like type of health facilities; need to get an appointment before visiting
PHC facility; waiting time to get check up by doctors and provision of medical supply were asked.
Around 75% of respondents reported to visit public healthcare facilities which were available at
every sub districts. About 18% of total respondents mentioned about the need to get an appointment
before visiting private clinics. None of the respondents stated about difficulty in getting appointment
as it could be easily done by telephone. In general availability of medicine did not seem to be a
problematic factor, as 97.5% of respondent mentioned that it was supplied by the public health
facilities itself or, if not, by a pharmacy near to the facilities. Among all these issues, respondents
expressed discontent only with the waiting time before getting check up especially in public health
clinics, puskesmas. This problem was remarkable as large percentage of respondent with 42% of total
were dissatisfied with long waiting time in health facilities. Hence, waiting time seemed to be an
important factor to be considered in availability in this study. As length of waiting time differed
depending on type of health facilities, for instance public or private healthcare facility, attributes
related to availability was studied separately for different facilities being used by respondents.
Type of primary healthcare facilities in use
Government healthcare centre, Puskesmas, was found to be the most common healthcare facility in
study area for primary healthcare. About 65% of total respondents from all three villages mentioned
puskesmas as the first PHC facility they visited. Total percentage of households visiting puskesmas
for PHC was about 53%, 79% and 60.5% for Kricak, Tegalpanggung and Tridadi respectively.
Private clinic was found to be the next common facility with 18% of total households visiting it;
mostly those belonged to higher socioeconomic class. Remaining 17% visited hospitals or sub-
puskesmas. The proportion of households visiting private clinics was greater in Kricak with about
22.5% followed by Tridadi with 20.5% and 10% in Tegalpanggung. This can be related to the result
of socioeconomic classification. As the portion of better off households was higher in Kricak, so is the
case with the visit to private clinic. Similarly, majority of households in Tegalpanggung belonged to
lower socioeconomic class so the proportion of visit to private clinics is least among three villages.
The proportion in Tridadi was in between Kricak and Tegalpanggung. However the difference is very
less with Kricak in both cases.
To check if socioeconomic status influenced the selection of PHC facility, relation between
socioeconomic class and type of facility visited was checked. Significant medium association was
found between these variables with Cramer‟s statistic .45 (p < .001). It was assumed that the
socioeconomic status influenced the selection of particular health facility to some extent. Hence the
perception on different dimensions of access based on type of facility being visited might change
accordingly.
46
Waiting time in PHC facility
Waiting time in PHC facility before getting check up by doctors showed a great range of difference
starting from 5 minutes to 3 hours. The range differed depending on the healthcare facility visited
which is shown in Table 5-2. Difference of around 50 minutes was found in the average waiting time
at puskesmas and private clinic. It was noted that the mean value for public facilities like hospitals,
puskesmas and sub puskesmas, was greater than the maximum waiting time for private clinics. The
overall mean values of waiting time for Kricak, Tegalpanggung and Tridadi are 50, 60 and 45
minutes respectively, which did not show much variation. Further subjective perception on the
waiting time was asked using a 5 point Likert scale starting from very short to very long. Figure 5-4
shows the percentage of overall and respondents‟ in village and per facility type that were
categorized in each level of Likert scale based on their response. Respondent‟s perception about
waiting time in health facilities was not very different when compared between villages. But large
difference was observed in case of private and public healthcare facilities. Subjective perception was
assumed to vary with personal as well as socioeconomic characteristics of respondents. For instance,
about 10% of respondents visiting a private clinic stated „long‟ for the waiting time of 20-30 minutes.
On other hand, about 25% of respondents visiting puskesmas stated „normal‟ for the waiting time
from 45-90 minutes.
% of
households
Waiting time in minutes
Facility visited for PHC Minimum Maximum Mean
Hospital 8.4 15 120 44
Puskesmas 65.2 10 180 65
Sub –puskesmas 8.7 10 90 38
Private Clinic 17.7 5 30 15
Overall waiting time in minutes before
doctor's check up without considering
different facility visited for PHC
100 5 180 52
Table 5-2: Descriptive statistics of waiting time in health facilities
Figure 5-4: Percentage of subjective perception on waiting time in
PHC
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Hospital Puskesmas Sub-
puskesmas
Private
clinic
Very long
Long
Normal
Short
Very short0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Hospital Puskesmas Sub-
puskesmas
Private
clinic
Very long
Long
Normal
Short
Very short
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very long
Long
Normal
Short
Very short
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
47
5.2.2 Descriptive Statistics for Accessibility to PHC
Physical accessibility was measured in terms of travel impedance. Objective and subjective questions
were asked regarding travel distance, travel time and mode of transport to reach the PHC facility. For
accuracy, geographic distance from household point to the reported public health facilities was
measured in GIS instead of using reported travel distance by respondents. Overall mean for measured
travel distance was about 1450 meters and travel time was about 10 minutes. Majority of respondents
around 70% stated the travel distance to be near and very near when asked in a 5point Likert scale.
Similar responses were found about travel time to reach the PHC facility. Only about 3% of total
respondent stated far for travel distance. Remaining answers were stated as normal. Figure 5-5 shows
the perception of respondents regarding travel distance and travel time for each sample village.
Regarding the mode of transportation, large portion of respondents (50.5%) used motorbike to visit
the healthcare centre and about 39.5% used bicycle or walk. As the physical distance was not a
problem, only a small portion (7%) of respondents used public transportation and remaining 3%,
those from well off class, travelled by car. Depending on respondent‟s opinion, it can be said that in
general physical accessibility did not seem to be a major problematic factor in access to PHC in the
study area.
5.2.3 Descriptive Statistics for Affordability of PHC
Affordability was measured including direct and indirect expense of healthcare service by asking
costs and opinion about various healthcare costs. Possession of health card issued by local
government, „Katu Sehat’ was asked as it was a relevant aspect in affordability that influences
respondents‟ subjective perception. About 55% of households stated to have the card, among which
54.5% are from lower socioeconomic class. Majority of households (50%) which did not have health
card belonged to the higher socioeconomic class. Primary healthcare service cost was found to be
nominal in all puskesmas even without the card. However, it differed and was higher for private
clinics. As private clinic was visited mostly by well of class, majority of total respondents stated
service cost to be inexpensive. When asked about travel cost, 93% of respondent stated it to be
normal or inexpensive. Similarly, 80% said medicine cost to be normal or inexpensive as in most
cases medicine cost was covered by the health card. After asking about individual costs, opinion on
the total PHC cost was asked. Only 13 % of total respondents expressed it as expensive. This could
be due to medicine cost in cases when respondents did not have health card or if medicine was not
Figure 5-5: Percentage of subjective perception on travel distance and travel time to PHC
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very far
Far
Normal
Near
Very near0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very far
Far
Normal
Near
Very near
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very long
Long
Normal
Short
Very short0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very long
Long
Normal
Short
Very short
48
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Hospital Puskesmas Sub-
puskesmas
Private clinic
Expensive
Normal
Inexpensive
Very inexpensive
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Hospital Puskesmas Sub-
puskesmas
Private clinic
Expensive
Normal
Inexpensive
Very inexpensive
provided free of cost by health facilities. 34% of total respondents said that total cost to be normal
and remaining 53% said it was inexpensive. The percentage of perceived attributes on total cost of
PHC per village and healthcare facility type is presented Figure 5-6. Although in general, people
have positive to neutral opinion about healthcare cost, in some cases up to 20% of respondents
expressed dissatisfaction towards it. It may not be a major issue but deserves attention.
Figure 5-6: Percentage of subjective perception on total cost for PHC
5.2.4 Descriptive Statistics for Acceptability and Adequacy of PHC
Acceptability and adequacy were measured from respondents‟ perception on issues related to cultural
or religious preference in selecting particular healthcare facility. Factors like preference towards
gender of medical personnel, physical appearance of facility, service opening hour, trust in medical
ability and opinion about inter-personal treatment from facility personnel were investigated. When
respondents were asked if choice given, will they have any religious preference in choosing particular
healthcare facility, only 12.5% of total respondents said yes. Further a question was asked about the
gender preference of doctors in PHC facility. At the first instance, majority of respondents answered
that gender of doctor was not an issue for healthcare. Question was then elaborated giving an option,
if there were equal number of male and female doctors in health facilities would they have any
preference to be treated by particular gender of doctor. Then 42.5% of respondent mentioned if
applicable they would prefer female doctors for female patients and only 9% said male doctors to
male patients. However, even after the given option, 48.5% of respondents still stated that gender of
doctors did not matter for both male and female patients. Figure 5-7 (b) displays the percentage of
two opinions on gender preference when applicable, by considering preference of male or female
doctors to respective patients as „yes‟ and no gender preference for both as „no‟. This result showed
that religious factor was not a prominent issue while selecting a healthcare facility though people had
certain preference towards the gender of doctors especially for female if given an option. However,
these issues did not seem to significantly influence peoples‟ acceptability of healthcare facility while
evaluating overall access to the facility at present context. Higher literacy rate could be a reason for
better state of acceptability as highly significant association was found between the level of literacy
and score of acceptability index (Cramer‟s V = .513, p < .001). Also the province was homogeneous
in terms of religion as 90% of population was Islamic. Such issues related to acceptability are
relevant in areas which are more heterogeneous in terms of religion, ethnicity or race.
While measuring adequacy, respondents did not complain about physical appearance or cleanliness of
the PHC facility, both public and private, as 99.3% of total respondent were satisfied with the
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
49
cleanliness of healthcare facility they visited. Opening hour of facilities also did not seem to be a
major issue, as 87% of respondent stated that the opening hour suits their working schedule.
When asked about the trust in medical ability of healthcare facility visited, 93% of respondent had
positive to neutral opinion and only 7% said it was bad. All respondents who were unsatisfied visited
puskesmas for healthcare. Majority (89.5%) of those who visit private clinic stated the medical
ability to be good and remaining expressed neutral perception. Chi square test was done to see if any
relation exists between the opinion in medical ability and the facility visited. Significant medium
association was found between these variables with Cramer‟s statistic of .41(p < .001). Although the
association was not very strong, it can be said that perception on medical ability, to some extent, was
based on the type of facilities people visited.
A question was asked about the feeling of inter-personal treatment by medical personnel. About 40%
of respondents expressed dissatisfaction and answered it to be bad, out of which 80% were visiting
puskesmas. 35% of respondents stated that the personal behaviour was normal and 25% said it was
good. Chi square test between facility type and opinion of personal treatment showed highly
significant medium association between these variables similar to the previous case of trust in medical
ability and type of facility (Cramer‟s V = .3, p < .001). Perceived opinions on different factors
related to acceptability and adequacy are shown in Figure 5-7.
From interviews with households and puskesmas staffs, it was known that there are four different
hierarchy of service provision by public health facilities; VIP, 1st class, 2nd class and 3rd class
depending upon quality of service and service cost that patients are willing to pay. Priority for service
was given starting from VIP to 3rd class. Patients with health card seeking for free PHC received the
3rd class service. When asked about difference in such services, respondents did not complain about
quality of service but mentioned about unfriendly behaviour of medical staffs.
(b) Gender preference for doctors
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
No
Yes0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
No
Yes
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
No
Yes
Tegalpanggung Kricak Tridadi
(a) Religious importance
(d) Trust on medical ability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Bad
Normal
Good
Very good
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Bad
Normal
Good
Very good
Tegalpanggung Kricak Tridadi
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Dirty
Normal
Clean
Very clean
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Dirty
Normal
Clean
Very clean
Tegalpanggung Kricak Tridadi
(c) Physical appearance of
PHC facility
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
No
Yes
Tegalpanggung Kricak Tridadi
50
Figure 5-7: Percentage of subjective perception on factors related to acceptability and adequacy
5.2.5 Overall Satisfaction Level with Access to PHC
After asking about the subjective opinion on each dimensions of access, overall satisfaction level with
access to PHC was also asked in 5 point Likert scale, considering all the factors in dimensions.
Majority of respondent with 48% of total expressed dissatisfaction or very dissatisfaction. 34% said
satisfied or very satisfied and remaining 18% expressed neutral opinion about the access to PHC
facility. The satisfaction level was then checked for different socioeconomic classes. Large portion of
households (45%) expressing satisfaction towards access were from higher socioeconomic class and
majority of households (51%) expressing dissatisfaction belonged to lower socioeconomic class.
When level of satisfaction was checked for different types of health facilities, remarkable difference
was found between public and private facilities. Around 60% of respondents visiting puskesmas
reported to be unsatisfied whereas this figure was less than 5% for private clinics. These results can
be interpreted with significant medium association between socioeconomic status and type of facility
being visited found by Chi square test in section 5.2.1. Figure 5-8 presents different level of
satisfaction within villages, socioeconomic cluster and per healthcare facilities. Significant strong
association was found between the type of facility (private and public) and the level of satisfaction to
access to PHC (Cramer‟s V = .51, p < .001).
Further to ensure the level of satisfaction, respondents were asked if their income is doubled do they
still want to visit the same PHC facility. 38% of respondent said they will visit the same facility and
62% said they want to change the facility. Chi square test was computed to see if there is any
association between the satisfaction level and willingness to change the health facility. Result showed
highly significant medium association between these attributes (Cramer‟s V = .42, p < .001). This
indicates that respondent‟s willingness to change the health facility was related to their satisfaction
level with access to healthcare facility and these attributes were related to each other.
(e) Inter-personal treatment by medical personnel
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very bad
Bad
Normal
Good
Very good
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Tegalpanggung Kricak Tridadi
Very bad
Bad
Normal
Good
Very good
Tegalpanggung Kricak Tridadi
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Hospital Puskesmas Sub puskesmas Private Clinic Hospital Puskesmas Sub-puskesmas Private clinic
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
51
Figure 5-8: Perceived satisfaction level with access to PHC
5.2.6 Influencing Factors to Overall Satisfaction Level
One of the research questions of this study was to find dominant factors that influence the evaluation
of perceived access to PHC. Answer to this question can be used to focus and prioritize such
dominant factors for further improvement of access. It was assumed that perception on those factors
had some relation with respondents‟ overall satisfaction level with access to PHC. During household
survey, overall satisfaction was asked after asking respondent‟s perception on each factor related to
dimensions of access. From earlier analysis in this section, majority of respondents expressed
dissatisfaction with waiting time in health facilities and inter-personal behavior by medical personnel.
All other factors related to dimensions of access to healthcare were found to be less problematic as
perceived by respondents. Hence, to see the relationship between factors related to each dimension of
access and the satisfaction level Spearman‟s correlation was computed between the subjective
opinion on different factors and overall satisfaction level. Aggregated 3 point Likert (satisfied;
normal; unsatisfied) was used, as it showed significant and stronger correlation between variables
than the 5 point Likert. Table 5-3 (a) shows the correlation matrix between factors under dimensions
of access and the satisfaction level when all samples were taken. Only few factors were found to have
statistically significant correlation between each other and with the final satisfaction level. Subjective
opinion on availability (waiting time in healthcare facility) and adequacy (inter-personal treatment)
showed comparatively strong positive correlation (r = .546 and r = .600 respectively at p < .001) with
the satisfaction level. Perception on medical ability of medical personnel also showed a significant
Tegalpanggung Kricak Tridadi
(a) Village
HSEC MSEC LSEC
(b) Socioeconomic class (c) Healthcare facilities
Private
clinic
Hospital Puskesmas Sub-
Puskesmas
52
relation with the overall satisfaction level, however with low correlation coefficient (r = .232,
p<.001).
Similar correlation was repeated for households visiting to different healthcare facilities to see if
other factors have any relation with the satisfaction level depending on the choice of facility type. For
public facility like puskesmas, almost similar result was found as earlier, which is shown in Table 5-3
(b). Satisfaction level of respondents was found to be influenced only by waiting time and inter-
personal treatment. There was no strong correlation between other factors except for travel distance
and travel time. Hence, waiting time and personal treatment was found to be dominant factors in
determining peoples‟ satisfaction with access to PHC in case of public health facilities.
Correlation computed for private clinics displayed different results than previous ones. Correlation
matrix for private clinics is shown in Table 5-3 (c). In case of private clinics, respondent‟s overall
satisfaction level showed significant positive correlation with their perception on other factors like
travel distance, cost and medical ability. Like in previous cases, satisfaction with waiting time and
personal treatment also had significant positive relation with the overall satisfaction.
Also, there was significant relation between factors like waiting time, medical ability and personal
treatment which was missing in case of puskesmas. As private clinics did not have provision of free
healthcare service unlike other public facilities, affordability was an important factor that influences
peoples‟ satisfaction level to some extent. However, the correlation was not very strong, as majority
of households visiting private clinics belonged to higher socioeconomic class who expressed the
service cost to be normal.
SO LOS WT TD TT TC CI GP PT PA MA
LOS 1
WT .546** 1
TD -.090 -.014 1
TT -.030 .048 .610** 1
TC -.018 .002 .136* .078 1
CI .106 .113 .093 .033 -.037 1
GP .008 .071 .043 .026 .050 .130* 1
PT .600** .225** -.126* -.068 -.032 .113 .021 1
PA .067 .050 .171** .117 .052 .050 .076 .015 1
MA .232** .067 -.080 -.038 -.173** -.011 -.103 .251** -.002 1
**. Correlation is significant at the 0.01 level (2-tailed)
*. Correlation is significant at the 0.05 level (2-tailed)
SO = subjective opinion, LOS = level of satisfaction with overall access to PHC, WT = waiting time in PHC facility, TD =
travel distance to PHC facility, TT = travel time to reach PHC facility, TC = total cost of PHC, CI = cultural importance in
choosing PHC facility, GP = gender preference of doctors, PT = personal treatment in PHC facility, PA = physical
appearance of PHC facility, MA = trust on medical ability of PHC facility
(a) Correlation matrix for total sample households
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
53
SO LOS WT TD TT TC CI GP PT PA MA
LOS 1
WT .501** 1
TD -.116 -.006 1
TT -.008 .111 .587** 1
TC .025 .058 .054 .016 1
CI .086 .128 .079 .046 -.092 1
GP -.062 .045 .068 -.022 .071 .156* 1
PT .552** .092 -.090 -.014 .047 .103 .042 1
PA .119 .113 .159* .210** .061 -.028 .133 .071 1
MA .046 -.071 -.038 -.040 -.100 -.014 -.168* .165* -.032 1
**. Correlation is significant at the 0.01 level (2-tailed)
*. Correlation is significant at the 0.05 level (2-tailed)
(b) Correlation matrix for samples visiting puskesmas
SO LOS WT TD TT TC CI GP PT PA MA
LOS 1
WT .695** 1
TD .411** .260 1
TT .019 -.126 .338* 1
TC .309* .292* .180 .080 1
CI .269 .215 -.007 -.157 .026 1
GP .241 .164 .154 -.012 .009 .258 1
PT .432** .367* .040 -.092 -.050 .387** .110 1
PA .110 .248 .135 -.091 .059 .312* -.013 .098 1
MA .529** .462** .368** .044 -.018 .380** .236 .290* .201 1
**. Correlation is significant at the 0.01 level (2-tailed)
*. Correlation is significant at the 0.05 level (2-tailed)
(c) Correlation matrix for sample visiting private clinics
Table 5-3: Correlation matrices between perception on factors of access and overall satisfaction
level with access to PHC
5.3 Synthesis of Indicators to Develop Summary Scores
Perceived satisfaction level of respondents across indicators was used to develop summary scores for
all dimensions. A total of 15 indicators measured in 5 point Likert scale were used in this process
which is shown in Figure 5-9. As acceptability and adequacy lacks objective indicators, subjective
perception about different factors related to access was used. While developing such summary scores
for all dimensions, equal weights were applied to each indicator. Distributions of summary scores for
each dimension are demonstrated by using box plot in Figure E-1 in Appendix E.
For easier interpretation, indicator values were standardized with ratio scale properties (standard
score, i = raw score i / maximum raw score). Though earlier analysis in section 5.2, showed
54
significant variation in the perception of respondents between indicators under same dimension, equal
weights were used for all indicators, based on the equal importance of all five dimensions in the
concept of access to healthcare. For instance, out of three indicators under availability, only waiting
time in health facilities was found to be the problematic issue in context of study area. However, ease
of getting an appointment to visit health facilities and provision of medicine supply cannot be less
prioritized while evaluating the state of availability related to overall access to healthcare. All these
factors might hold equal importance in evaluation of access at wider perspective of policy making.
Figure 5-9: Synthesising indicators to develop summary scores for dimensions of access to PHC
Perceived state of access
to PHC viewed from
different dimensions
Perceived satisfaction
level with:
Accessibility
0.5
0.5
- Travel distance to PHC
- Travel time to reach PHC
Availability
-Ease to get appointment
-Medicine availability
- Waiting time in PHC
0.33
0.34
0.33
Affordability
-Travel cost to reach PHC
-Service cost of PHC
- Medicine cost
-Total cost of PHC
0.25
0.25
0.25
0.25
Acceptability
-Religious aspect
-Gender preference of
medical personnel
0.5
0.5
Adequacy
-PHC service opening hour
-Physical appearance of
PHC facility
-Medical ability of medical
staffs
-Inter-personal treatment
from medical staffs
0.25
0.25
0.25
0.25
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
55
Average summary score was calculated by dividing the sum of total scores with the number of cases
for each villages, health facilities and socioeconomic classes. This score was used as the
representative value for respective cases. Average summary scores of dimensions for villages and
healthcare facilities are plotted in a radar chart shown in Figure 5-10. These scores are listed in Table
E-1 in Appendix E.
Accessibility scores are highest for villages Tegalpanggung and Kricak as compared to other
dimensions within villages. As Tridadi belongs to rural regency, geographic distance was
comparatively larger than other two villages. Also considerable number of respondents from Tridadi
reported to visit health facility in City of Yogyakarta rather than near facilities within their sub
district. Affordability and acceptability also scored higher values as compared to availability and
adequacy in all three villages. These scores somehow refers to the result obtained in section 5.2 where
only waiting time in health facilities and inter-personal treatment from medical personnel were found
to be a concern of respondents‟ dissatisfaction. There was small difference in scores of each
dimension between villages, which indicates similarity in state of access in villages. However
satisfaction level of respondents on different dimensions of access within each village showed
remarkable variation, for instance difference in accessibility score with availability and adequacy.
Remarkable variation can be seen in the scores when different types of health facilities used by
respondents were compared. Figure 5-10 (b) gives a clear interpretation of how perceived satisfaction
level of respondents varied based on the type of health facilities they visited. The score of availability
and adequacy are significantly higher for private clinics as compared to other public facilities. As the
satisfaction level with inter-personal treatment and the quality of service of private clinics was
significantly higher than that of other public facilities. And these scores are lowest for puskesmas. But
in case of affordability, there is swap in the scores. It is very low for private clinics when compared
between scores of other dimensions within same facility. Also the score is lowest when compared with
affordability scores of other facilities, whereas it is much higher for puskesmas and sub puskesmas.
The reason behind can be related with the findings of section 5.2. As public health facilities provides
free primary healthcare and medicine to patients with health card and majority of patients visiting
puskesmas, for example, have the health card. Hence, respondents visiting public health facilities
expressed higher satisfaction towards affordability. Although majority of respondents visiting private
clinics were from higher socioeconomic class, they reported the cost of service and medicine to be
relatively expensive. As the service is free or in nominal cost, majority of population visits
puskesmas. This results in a long waiting queue for check up in the facility unlike in private clinics.
Sub puskesmas has the highest score for accessibility and that of hospital is lowest. These figures well
fit the actual context as sub puskesmas are the supporting health facilities allocated per sub districts.
There are normally two to three sub puskesmas per sub districts, whereas hospitals are the biggest
health facilities in terms of service capacity and provision and are allocated per regencies. So the
difference in geographic distance to these facilities was large. Acceptability had relatively higher
scores with small variation in both cases of villages and facilities. As respondents expressed that
religious factors and gender issues were not relevant matters of concern in choosing health facilities
and they were satisfied with the existing situation in terms of acceptability.
56
Figure 5-11 shows the standardized residuals of percentage of respondents‟ perceived satisfaction
with each indicator of access. Positive residuals imply that frequency of satisfied respondent is higher
than expected average and negative residual means the reverse in all cases.
Frequency of respondents, satisfied with both indicators of accessibility was lower from village
Tridadi as compares to other two, showed in Figure 5-11 (a). In availability, frequency of satisfaction
with waiting time in facility was distinctly low in Tegalpanggung, whereas it was higher for Kricak.
Satisfaction with indicators of affordability was higher in Tegalpanggung. Tridadi had low
satisfaction with travel cost which can be seen in relation with lower satisfaction with travel distance
and time. Difference was less for indicators of acceptability, as respondents‟ perception was more
homogeneous in these factors. Variation between villages was observed in case of adequacy.
Comparatively Tridadi was in better satisfaction level with medical ability and inter-personal
treatment of medical personnel. And, inter-personal treatment was a notable problem with
respondents in Tegalpanggung.
Hospital
Puskesmas
Sub puskesmas
Private clinic
(b) Healthcare facilities
(a) Villages
Tegalpanggung
Kricak
Tridadi
Figure 5-10: Summary score chart for dimensions of access to PHC
For: (a) sample villages and (b) different healthcare facilities
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
57
More distinct variation was observed in case of different health facilities displayed in Figure 5-11
(b). It was clear that when compared, large portion of respondents expressing problem with physical
distance were the one visiting hospitals. Significant difference was seen in case of waiting time in
facility; healthcare service cost; total cost; trust in medical ability and personal behaviour between
puskesmas and private clinics. Among these indicators, only healthcare cost was the matter of less
satisfaction to respondents visiting private clinics when compared with the cost of puskesmas.
Satisfaction with waiting time, and indicators under adequacy was much higher in case of private
clinics and was low for puskesmas.
Average summary scores were used to see variation in dimensions of access across different
socioeconomic classes to observe homogeneity within villages in terms of access to PHC. Table 5-4
presents the average summary scores of dimensions for each socioeconomic class per village. There
was no distinct variation in individual dimension scores between socioeconomic classes in all three
villages. However, some variation was found between scores of different dimensions within each
socioeconomic class in Tegalpanggung and Kricak. All socioeconomic classes in these two villages
Figure 5-11: Standardized residuals for perceived satisfaction with various factors of access
Perceived satisfaction with:
1.a = Travel distance to PHC
1.b = Travel time to reach PHC
2.a = Ease to get appointment
2.b = Medicine availability
2.c = Waiting time in health
facility
3.a = Travel cost
3.b = Service cost
3.c = Medicine cost
3.d = Total cost
4.a = Religious aspect
4.b = Gender preference of
medical personnel
5.a = Service opening hour
5.b = Physical appearance of
facility
5.c = Trust in medical ability
5.d = Inter-personal treatment
(a) Villages
(b) Health facilities
58
had higher score for accessibility and lower for adequacy. Scores for affordability and acceptability
were comparatively high than availability and adequacy with small variation. Higher socioeconomic
classes in all villages had slightly higher scores for adequacy and availability than middle and lower
socioeconomic classes, whereas it had lower scores in affordability. This can be related with the type
of facility visited by different socioeconomic class, as large percentage of respondents visiting private
clinics falls in higher socioeconomic class as discussed in section 5.2.1. Tridadi was most
homogeneous in terms of variation in dimensions of access to PHC within and also between all
socioeconomic classes. As very small variation was obtained between scores of different
socioeconomic classes, radar chart could not display clear interpretation, hence it was not used.
Accessibility Availability Affordability Acceptability Adequacy
Tegalpanggung
HSEC
MSEC
LSEC
0.81
0.82
0.79
0.73
0.70
0.71
0.74
0.79
0.78
0.77
0.76
0.76
0.71
0.68
0.66
Kricak
HSEC
MSEC
LSEC
0.81
0.82
0.8
0.77
0.71
0.73
0.70
0.78
0.81
0.78
0.76
0.75
0.70
0.69
0.67
Tridadi
HSEC
MSEC
LSEC
0.72
0.73
0.73
0.76
0.70
0.71
0.74
0.76
0.75
0.76
0.77
0.76
0.75
0.72
0.70
Table 5-4: Summary score for dimensions of access per socioeconomic class
The reason for this homogeneity was due to the fact that majority of households (74%) from all
villages visited public health facilities like hospitals and puskesmas for primary healthcare. People
opt for public facilities, primarily puskesmas, because of short geographic distance; ease in medicine
availability and satisfactory quality of service in terms of medical ability and cleanliness, in addition
to nominal service cost. Hence there was similarity in perception over factors related to access from
respondents regardless of their socioeconomic status. As the number of respondents from high
socioeconomic class visiting private clinics was low (18%) as compared to the number visiting
puskesmas from same socioeconomic class, their perception was over shaded by the large number of
respondents visiting public facilities while taking average score. Variation observed for different
health facilities in Figure 5-10 (b) can also explain the variation between socioeconomic classes to
some extent. As around 80% of respondents visiting private clinics are from higher socioeconomic
class and only 19% from middle socioeconomic class, the summary scores of private clinics also
reflects the status of access for higher socioeconomic class. In this manner, it can be said that some
variation do exists between different socioeconomic classes.
5.4 Existing Situation of Access to PHC from Policy Perspective
To address the research questions of second sub objective, access to health care was analyzed as per
the official norms mentioned in existing health policy in the province of Yogyakarta (DIY). The
health policy in DIY primarily focuses on physical accessibility, availability of public health facilities
and affordability for healthcare service. Information on geographic distances to public health
facilities and total number of doctors per village, obtained from census 2005 was used for analyses.
Source of all health policies and standards used in this study was Statistics of Indonesia obtained from
Note: High values indicate high level of access; 1 = maximum value
[HSEC = Higher socioeconomic class; MSEC = Middle socioeconomic class;
LSEC = Lower socioeconomic class]
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
59
the Gadjah Mada University (Pustral UGM). As per the norm, there should be at least one puskesmas
headed by a medical doctor at every sub district level supported by 2 to 3 sub puskesmas headed by
medical nurses. This is related with accessibility as well as availability. It addresses the issue of
geographic distance to healthcare considering administrative boundary and availability by
considering puskesmas to population ratio per sub district. This norm was well met at existing
situation, as all 78 sub districts in DIY contained at least one puskesmas and supporting sub
puskesmas. Considering geographical distance, 70.5% of total population in the province (271
villages out of 438) was within 2 kilometres and 88% population (327 villages) within 4 kilometres,
assuming 4 km per hour as normal walking distance. All 45 villages in the city of Yogyakarta were
within the distance of 2 km from nearest puskesmas in respective sub districts covering 15.5% of total
population in the province. In terms of distance to hospital, 44.5% of total population (138 villages),
including all in city of Yogyakarta, in the province was within 4 km distance from hospital. Variation
in geographic distance to puskesmas and hospitals between villages in DIY is shown in Figure 5-12
(a) and (b) respectively. This can be related with the finding of previous analysis for accessibility in
section 5.2.2, where majority of respondents in sample villages reported physical accessibility as less
problematic issues in access to healthcare. Regency Gunung Kidul had the highest area and lowest
population density among all five in the province. Also most of villages, except those located in
central area in this regency had primarily agricultural landuse. As settlement was concentrated in the
central part of the regency, health facilities were also located accordingly. Hence, looking at Figure
5-12 (b), large number of villages in regency Gunung Kidul showed higher travel distance to hospital
except for villages located in central part of the regency. Villages at larger geographic distance from
hospitals can be related with lower population density in Figure3-2 and vice verse.
According to the health policy in DIY, ratio of puskesmas to population was 1: 120,000 and that for
hospital was 1:240,000. The ratio of puskesmas to population in all sub districts was distinctly
smaller than the standard threshold as none of the sub district had population this large. Largest
population of sub district in DIY was 117,130 in Sleman regency. Hospital to population ratio per
regency was also found smaller than the standard, except for regency Gunung Kidul which was
around 1: 250,000. Further to elaborate the analysis of availability, population to doctor ratio was
checked per sub districts using the number of doctors working at public health facilities and
population data per village in census. Analysis was done at sub district level as it was the smallest
administrative unit for the allocation of puskesmas according to health policy. Dash (2000) stated
that average doctor population ratio in Indonesia was around 1:7,000 according to the World
development report. Hence due to lack of specific standard threshold for doctor population ratio, this
ratio was taken as the benchmark for the analysis in DIY. Variation in ratio between sub-districts is
shown in Figure 5-13 (a). Out of 78 sub districts 43 had the ratio below 1: 7,000 which includes all
sub districts within city of Yogyakarta. 14 sub districts had the ratio 1:7,000 to 1:14,000 and other
14 sub districts had the ratio larger than 1: 14,000. Majority of sub districts with higher ratio was in
regency Gunung Kidul and Kulon Progo. All three sub districts containing sample villages had the
ratio below 1:7,000. Figure 5-13 (b) shows the relative variation in population to doctor ratio within
the province of Yogyakarta using standard deviation. Average ratio in the province was found to be
1:9900, with standard deviation of 12600, which was larger than the average ratio in Indonesia. Huge
variation was observed in the population doctor ratio starting from 1: 425 in regency Sleman going up
to 1: 55,000 in Gunung Kidul. Comparatively city of Yogyakarta and regency Sleman were in better
situation in terms of availability of medical doctors. This result was contradicting with the higher
dissatisfaction of respondents with long waiting time in public health facilities reported during
household survey. There was a huge difference in population doctor ratio when compared with the
average ratio of neighbouring country like Singapore.
60
Dash (2000) stated that population to doctor ratio in Singapore was 1:900. Further comparing it with
developed countries like Japan (1:600) and United State (1:400), the average population-doctor ratio
in Indonesia was found to be very large. Also when referred to the estimated ratio of physicians to
population by human resources for health, World Health Organization (WHO), the existing ratio was
found to be very large. WHO stated that although there is no fix standard of ratio of population to
medical service personnel, countries with less than 25 medical staffs including physicians and nurses
per 10,000 populations cannot achieve adequate service coverage for primary healthcare. ESCAP
(2009) stated the estimated ratio of 1 physician to 1000 population in context of Asia in WHO
statistical information system 2008. When compared with this figure, only 4 sub districts; 2 in Sleman
and 2 in City of Yogyakarta, were found to have meet the ratio estimated. This could be related to
the respondents‟ complaints about long waiting time in health facilities.
(a) (b)
Distance to puskesmas (km) Distance to hospital (km)
(a) (b)
Figure 5-12: Geographic distance to health facilities per village in DIY,
Data source: Census 2005
Figure 5-13: Population-doctor ratio per sub districts in DIY Data source: Census 2005
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
61
One of the major health policies in the province of Yogyakarta focuses in the provision of nominal
primary healthcare service in almost all public health facilities. As explained earlier in section 3.2.2,
better service quality, accessibility and affordability are the part of „Healthy Indonesia 2010‟ which
is one of the policy visions in line. The policy concerned to affordability was primarily focused to
facilitate poor population by providing government health insurance card, commonly known as a
health card. This policy can be considered to be well implemented as more than 55% of sample
households in Tegalpanggung and Tridadi were getting free primary healthcare service including
medicine provision with the possession of health card. This figure was around 45% for Kricak.
Among total sample, 80% of households having a letter of poor reported to have a health card, which
was a criteria for getting the health card according to the norm. Even without health card, service cost
for primary healthcare in all public health facilities was found to be nominal as 87% of total
respondents were satisfied with the total cost of healthcare. Although issue of service quality
appeared in health policy, there was not any standard threshold to evaluate the existing quality of
health service. Hence it was done based on respondents‟ perception from household survey. Quality
of service basically means peoples‟ trust on the medical ability of health facilities. When asked, only
7% of respondents expressed dissatisfaction with the medical ability in terms of quality of service. As
medical service is given by or under supervision of professional medical doctors at puskesmas, people
expressed their trust over the service quality. Other issues related to the service quality like
cleanliness of facilities, availability of equipment and laboratory service for primary healthcare were
also found to be considerably in satisfactory level.
5.5 Scale Effect in Analyzing Socioeconomic Attributes and Access to PHC
Common variables between census and primary data were selected considering their availability and
relation to socioeconomic characteristics and access to healthcare for comparative analyses between
aggregated and disaggregated data. These variables are explained in Table 5-5.
5.5.1 Comparing Individual Variables from Census and Primary data
As a preliminary analysis, information on individual variables from census 2005 was compared with
that from primary household data. Significant difference was found in the information about
socioeconomic characteristics like possession of letter of poor and health card. These variables were
important as they reflect socioeconomic status of people and also from affordability aspect of access
to primary healthcare. Census presented very low percentage (less than 2%) of households with the
letter of poor in all three villages. This figure significantly differed from the finding of primary data,
which showed around 55% of households with the letter of poor in Kricak and more than 70% in
Tegalpanggung and Tridadi. Similarly, according to census, public piped water supply was the
dominant source of water in Kricak and Tridadi. But from field survey, it was found that around 60%
households in both villages were using well water, mostly shared community well. Regarding type of
toilet in all villages, census information was a good representation, as more than 70% of households
have private toilet which was mentioned as the dominant toilet type in census. Observation over the
percentage of households possessing health card for primary healthcare also showed large difference
of 72% in Tegalpanggung and 45% in Tridadi, as census present low figures for this attribute. But in
case of Kricak, census presented higher portion of 89.5% households with the health card, which was
found to be 67.5% from household data. In contrary, census showed low portion of households with
telephone connection, 14.5% in Kricak and less than 5% in other two villages. Collected primary
data showed more than 55% of households with telephone connection in all three villages.
Aggregated data on access to healthcare was available for accessibility in terms of geographic
distance and perception on ease to reach healthcare facilities. Hence those attributes were selected for
62
comparative analysis. Variation was found in physical accessibility to healthcare when check at
different level of administrative unit. Measured distance between sample households and healthcare
facilities (hospital and puskesmas) being visited by respective households was used to compare with
census figure. Distance to hospital obtained from primary source was found to be greater in
Tegalpanggung and Tridadi. Difference was much higher for Tegalpanggung. In case of Kricak,
census showed a smaller distance than the average value of primary data. Distance to puskesmas from
villages also showed some variation.
Table 5-5: Common variables in census 2005 and household data
Table 5-6 presents the descriptive statistics of physical distances to these healthcare facilities
obtained from primary data. Although there was not large difference in the average value of measured
distance and census, except in case of distance to hospital from Tegalpanggung, large variation in
minimum and maximum range of distance was observed in most of the cases. Around 40% of
households in Tegalpanggung and Tridadi did not visit hospital for primary care. Remarkable
variation in census figure and primary data was found in Tegalpanggung. Distance to hospital from
this village was 500 meters in census, whereas none of the sample household was within this distance.
The minimum distance from primary source was 1000 m and 15% of sample households are travelling
3000 to 4000 meters to visit a hospital. Also in Tridadi, only 24% of sample was within the distance
mentioned in census (4000 m) and about 31% was between a distances of 6000 to 11000 meters.
Similarly, the distance to puskesmas in census was not found to be a good representative for
Tegalpanggung and Kricak, as large percentage of sample was travelling greater distance than as
mentioned in census. This analysis showed the inconsistency in the results obtained from primary data
source, compared to the aggregated census figures. On the other hand, census figure was found to be a
good representative of distance to hospital in case of Kricak and distance to puskesmas for Tridadi as
more than 60% of sample fell within the distance as mentioned. Figure 5-14 shows the measured
Euclidean distance from each sample households to the visited hospitals and puskesmas as reported.
This displays the variation in travel distance to health facilities within each village, prominently for
hospitals from rural village Tridadi.
Attributes related to
socioeconomic status
Description according to Census
Letter of poor
Percentage of households within village possessing the letter of poor (see
section 3.2.2 for description)
Toilet type
Dominant toilet type in village (shared public toilet; private toilet and no toilet)
being used by majority of household in the village
Water source
Dominant source of water supply in village. (River water, ground water – hand
pump, well and public piped water supply)
Telephone connection Percentage of households within village with telephone connection
Attributes related to
access to healthcare
Distance to hospital Average physical distance from village centre to nearest hospitals
Opinion about distance to
hospital
Subjective opinion about ease to reach the nearest hospital based on physical
distance in four scale; very easy, easy, difficult and very difficult
Distance to puskesmas
Average physical distance from village centre to nearest puskesmas usually
located in same sub-district
Opinion about distance to
puskesmas
Subjective opinion about ease to reach the nearest puskesmas based on physical
distance in four scale; very easy, easy, difficult and very difficult
Possession of health card Percentage of households possessing government health card
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
63
Based on the nearest geographic distance, census stated a subjective opinion on the ease in reaching
those health facilities. As distance was not considered to be a problem in case of these villages, all of
them were given „easy‟ or „very easy‟ status in terms of physical accessibility in census. The
perception over physical distance obtained from primary data also stated accessibility to be less
problematic, in spite of the fact that people does not necessarily visit the nearest healthcare facilities
as per assumption. However some variation exists in the perception. Chi square test was done to see if
there is any association between the respondent‟s perception over accessibility measures and the
mode of transport they use to reach health facilities, which was asked during survey. Significant
medium relation was found between these attributes with Cramer‟s statistics of .46 (p < .05). Hence,
there was large variation in the respondents‟ perception over travel distance.
Due to lack of additional information on other dimensions of access to healthcare in census data, this
study was limited to compare variables related to accessibility, although it did not seem to be a
problematic issue in access to PHC in study area.
Figure 5-14: Euclidean distance from households to heath facilities
Tridadi
Kricak
Tegalpanggung
Tridadi
Kricak
Tegalpanggu
ng
64
Comparing census distance to hospitals with primary data (all distance is in meters)
Tegalpanggung Kricak Tridadi
Distance in census = 500 m Distance in census = 3000 m Distance in census = 4000 m
Primary data
% of households within distance
Less or equal to 1000 = 0% Less or equal to 3000 = 64% Less or equal to 4000 = 24%
1001 - 2000 = 35% 3001 - 6000 = 0% 4001- 6000 = 5%
2001 -3000 = 9% 6001 -7000 = 3% 6001 -7000 = 15%
3001 - 4000 = 15% greater than 7000 = 16%
Do not visit hospital = 41% Do not visit hospital = 33% Do not visit hospital = 40%
Mean = 2160 Mean = 2255 Mean = 5215
Minimum = 1000 Minimum = 1225 Minimum = 1050
Maximum = 4045 Maximum = 7200 Maximum = 11000
Comparing census distance to puskesmas with primary data
Distance in census = 500 m Distance in census = 500 m Distance in census = 3000 m
Primary data
% of households with distance
Less or equal to 500 = 18% Less or equal to 500 = 30% Less or equal to 3000 = 60%
501 – 1000 = 44% 501 – 1000 = 53% 3001 – 6000 = 3%
Greater than 1000 = 17% 1001 – 3000 = 7% Greater than 6000 = 7%
Do not visit puskesmas = 21% Do not visit puskesmas = 28% Do not visit puskesmas = 30%
Mean = 750 Mean = 625 Mean = 1970
Minimum = 240 Minimum = 1210 Minimum = 500
Maximum = 1210 Maximum = 3000 Maximum = 7850
Table 5-6: Distance to healthcare facilities from Census and primary data
5.5.2 Comparing Variation within Village with Variation at Sub district Level
Coefficient of variation (CV) was computed to further compare variability in selected socioeconomic
and access attributes at village level with that at sub districts level, using primary and secondary data
respectively. Variation in common socioeconomic attributes was found higher for all villages as
compared to their sub districts. This is shown in Table 5-7 (a). Although, private toilet was the
dominant toilet type in all sample villages, some variation was found within each village, whereas
census had a single value of toilet type for all three villages. Significant variation was observed in the
possession of health card in village Kricak and Tridadi which is directly related to affordability in
access to PHC.
Variation, in terms of access to healthcare, was found in physical distance to hospitals and puskesmas
at village level in Tegalpanggung, shown in Table 5-7 (b). This variation could not be observed at its
sub district level, as all villages in this sub district were given same value for respective distance
according to census. The variation in physical distance to puskesmas was significantly large in
Tridadi when compared with the variation at its sub district using aggregated village data. As Tridadi
belongs to rural regency, significant portion of respondent reported to visit health facility in
Yogyakarta city, resulting in larger variation in travel distance as well as their perception over it. In
case of Kricak, variation was found to be less than the variation in its sub district, however the
difference was small. Regarding the relative subjective perception on these distances, sub districts of
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
65
Tegalpanggung and Kricak does not show any variation as all the villages within these sub districts
were stated as having a „very easy‟ accessibility to mentioned facilities in census.
Table below displays the variation at sub district level using census data available for each village
within the sub districts. Rows specified as „census(s)‟ shows this variation. Rows specified as
„Primary data (v)‟ displays the variation within sample villages using primary household data.
Purpose is to compare variation within village to that with its corresponding sub district.
Tegalanggung belonged to sub district Danurejan; Kricak to Tegalrajo and Tridadi to Sleman, hence
comparison is done correspondingly.
(a) Socioeconomic variability in terms of coefficient variation (CV)
Data source
Sub-district (s)
/ Village (v)
Type of
toilet
Water
source
Telephone
connection
Possession of
letter of poor
Possession of
health card
Census (s) Danurejan NA NA 0.72 0.45 0.54
Primary data (v) Tegalpanggung 0.28 0.60 1.03 0.60 0.70
Census (s) Tegalrajo NA 0.16 1.18 0.39 0.13
Primary data (v) Kricak 0.24 0.54 1.50 0.86 1.12
Census (s) Sleman NA 0.16 0.8 0.64 0.22
Primary data (v) Tridadi 0.22 0.48 1.01 0.65 0.96
(b) Variability in accessibility to PHC in terms of coefficient variation (CV)
Data source
Sub-district (s) /
Village (v)
Distance
to hospital
Subjective
perception on
distance to hospital
Distance to
puskesmas
Subjective
perception on
distance to puskesmas
Census (s) Danurejan NA NA NA NA
Primary data (v) Tegalpanggung 0.435 0.276 0.375 0.279
Census (s) Tegalrajo 0.556 NA 0.707 NA
Primary data (v) Kricak 0.450 0.264 0.675 0.288
Census (s) Sleman 0.520 0.248 0.476 0.248
Primary data (v) Tridadi 0.530 0.261 0.941 0.257
Table 5-7: Socioeconomic and access to PHC variability in terms of CV
[NA = not applicable; if census have one value for all villages within that sub district]
66
6. Discussions on Findings
This chapter presents the analytical discussion of the results obtained to address sub objectives of
this study in three main sections. Main findings and limitations in methodologies applied in this
research are discussed. In the first section, results of variation in access between villages and
different socioeconomic classes are discussed. Perceived major factors that influences the overall
satisfaction level are also discussed in this section. Second section discusses the result of access to
healthcare in relation to contextual heath policies comparing it with results at micro level. Last
section contains a discussion over the findings of variation in results obtained from aggregated
and disaggregated data source.
6.1 Sub-objective 1: Measuring Access to PHC at Micro Level
Socioeconomic stratification
To address one of the research questions of sub objective 1, related to variation in access across
different socioeconomic strata, sampled households were classified into three classes using two step
cluster analysis. This stratification was relevant to measure intra village variation in terms of
socioeconomic characteristics which was assumed to have direct relation with access to healthcare.
Analytical discussion over the strength and limitations of the method used for stratification is
important as large part of subsequent analyses and findings was related to this result. Along with
advantages offered by cluster analysis explained in section 4.6.1, it had some limitations which were
experienced in this study. As explained by Wong (2006), cluster analysis require detailed and often
debatable operational decisions throughout statistical procedure. First, measurement of some form of
similarity in attributes between samples is needed in order to decide number of clusters. Once
different clusters were obtained, next step was the profiling of clusters to give them appropriate
labelling based on the composition of various input attributes. Although clusters were labelled
according to the characteristics exhibited by them, it was somehow a subjective judgement. For
instance, middle socioeconomic cluster in this study did not exhibit terribly distinctive characteristic,
rather it had some common characteristics from both higher and lower socioeconomic clusters. This
could be one reason that summary scores for dimensions of access developed for each socioeconomic
class within 3 sample villages in section 5.3 (Table 5-4) did not show much variation. Also other
statistical analyses, for example, Chi square did not have strong association (Cramer‟s V = .45)
between these classes and other variables like selecting type of health facilities. Another limitation is
that no ranking of individual cases within each cluster can be obtained as the samples were either in
or out of the cluster. Other cluster analysis such as K mean offers relative ranking of cases within
each group based on how strongly cases represent the belonging cluster, this was lacking in two step
cluster analysis. However, due to the advantage of simplicity and mainly its ability to create clusters
using both categorical and continuous variable two step cluster analysis was used for this study.
One of the assumptions made for selection of sample village was socioeconomic heterogeneity within
villages. This assumption was met as each sample village was composed of considerable percentage
of high and low socioeconomic classes. Also the assumption that poor informal settlement exists along
river bank, was met in case of village Tegalpanggung (see Figure 5-3).
Measuring access to PHC
Factors related to physical accessibility and availability of health facilities has been important
concerns in most of the empirical studies in evaluating access to healthcare. Literatures like Black,
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
67
Ebener et al. (2004); Bagheri, Benwell et al. (2005); Amer (2007) etc have been able to demonstrate
significant variation in these spatial component of access while evaluating health service variation.
However these dimensions in general, were not found to be the major problematic issues in this study
as the result of high priority given in the implementation of existing health policies related to physical
accessibility and availability of public healthcare facilities. Below 10% of total respondents,
dominantly from rural regency Sleman, stated physical distance to hospitals or puskesmas to be far.
This was not due to absence of health facilities within their administrative boundary, rather due to
user‟s personal preference to visit facilities in city of Yogyakarta.
Along with above mentioned spatial components, affordability and acceptability has also gained
priorities in evaluating equitable access to healthcare referring to literatures like Andersen,
McCutcheon et al. (1983); Fosu (1989); Obrist, Iteba et al. (2007) etc. However, as the result of free
or nominal cost of primary healthcare service including medicine supply in public facilities,
affordability in general was perceived as less problematic issue in the study area. Only about 13% of
total respondents were discontent with the total cost for primary healthcare. Noticeable fact related to
affordability was that none of the respondents from low socioeconomic class visited private clinic in
spite of their feeling about better service quality, short waiting time and good inter personal treatment
being provided by private clinics. The reason was directly linked with the higher service cost in
private facilities. This was confirmed when majority of respondents visiting public hospitals or
puskesmas stated that they would like to visit private facilities for primary healthcare, if their income
gets doubled. Also the satisfaction level with total cost showed a negative correlation with the
satisfaction level with trust in medical ability of health facilities (see Table 5-3. a). As respondents
visiting private clinics stated the cost to be higher, however they were very satisfied with medical
ability of the facility. Therefore affordability still had significant relevance in access in this manner.
Long waiting time under availability and unfriendly inter-personal treatment by medical staffs under
adequacy, were found to be dominant factors influencing the overall satisfaction of respondents
especially visiting public health facilities. Only satisfaction with these two factors showed highly
significant positive correlation with the overall satisfaction level (r = .546; r = .600 respectively at p
< .001). All other factors under five dimensions of access did not seem to influence the overall
satisfaction level of public facility users. Hence, it can be assumed that people tend to emphasize on
negative issues when asked about their satisfaction with present situation of access to PHC.
Satisfaction with factors like travel distance, service cost and trust in medical ability were important
in case of private health facility users, as these perceptions were found to influence the overall
satisfaction of people going to private clinics. Noticeable difference was observed in the perception
about similar waiting time between public and private facility users. As puskesmas provided free
primary healthcare for poor households, people might have accepted the longer waiting time or they
might be used to this situation hence, found it normal in general case. But majority of households
visiting private clinics belonged to higher socioeconomic class, who might be willing to pay more but
expects faster service. Due to the hierarchy in service provision by public health facilities, patients
seeking for free service might have experienced unfriendly behaviour by medical personnel. This
might create an issue of inequality in terms of inter-personal treatment among different
socioeconomic strata.
To evaluate combined effect of relevant factors under each dimension of access, summary scores were
developed which showed some limitations in this study. Along with the advantage of easy
interpretation especially at decision making level, a common disadvantage of developing summary
scores was that it loses underlying information to some extent. For instance, the problem of waiting
time and inter-personal behaviour in public health facilities was over shaded by the higher score of
other corresponding indicators while developing availability and adequacy summary scores. Another
68
limitation of this process was the weighting system. Although equal weights were applied to each
indicator, the relative value differed due to different numbers of indicators under different
dimensions. Larger the number of indicators within a dimension, lesser will be the impact while
developing a summary score for dimension. This influenced the final score for different dimensions
with varying number of indicators. This issue was addressed by using standardized residuals which
helped to visualize the variation in perceived satisfaction on each indicator under dimensions of
access. This also enabled to observe indicator(s) with large variation in perception within each
dimension, which could be hidden while using summary scores.
From analyses, intra village variation in access to PHC was more prominent than inter village.
Presence of different socioeconomic classes within each village met the assumption about intra
village heterogeneity which to large extent influenced the selection of health care facilities, public or
private. Further the perceived satisfaction with individual factors as well as overall access was greatly
related to the type of facility being visited.
6.2 Sub-objective 2: Existing State of Access in Relation to Health Policy
From the results obtained in section 5.2, existing situation of access to healthcare in study area can be
considered to be in accordance to the current health policy standards of DIY. Physical distances to
healthcare, availability of public health facilities to provide affordable and adequate quality of
service were the main health policy visions in line. Traditional approach in evaluating access to
public services has also given physical accessibility and availability of public facilities vital
importance while measuring service deprivation. These dimensions have also received considerable
attention in the policy aspect of Indonesia at various hierarchies starting from national level, Ministry
of Health, to district level health organizations. All sub districts in DIY contains one or more public
healthcare centres, puskesmas, which was one of the criteria in health policy. Also the ratio of
puskesmas to population as stated in policy was met in present condition. However, there was no
consideration of geographic area of sub districts for allocating health facilities, as there was vast
difference in area and population in sub districts in DIY. Sub districts in city of Yogyakarta were
small with minimum area of 0.65 km2 but highly pupulated and those in rural regencies were large
with maximum area up to 105 km2 with less population in Gugung Kidul. Therefore, geographic
distance to health facilities differs accordingly.
Two sample villages in this study were from urban sub districts in the city of Yogyakarta with areas of
1km2 and 3 km2. The third village was in periphery of the city in Sleman regency with 31 km2 area.
Hence physical accessibility did not seem to be a problematic factor in access to PHC except in few
cases from Sleman sub district. However, situation might be different in larger rural sub districts with
sparse population settlement, although it fulfils the policy standard.
The ratio of puskesmas to population per sub district in the whole province was within the policy
standard which was 1:120,000. But waiting time in puskesmas under availability was found to be one
of the main problems as perceived from household survey also from interviews with medical staffs in
five puskesmas. This refers that the ratio stated in policy norm is large which does not reflect reality,
hence require further attention. Also there was no consideration of ratio of medical staffs to
population in policy standard. Ratio of population to doctor in sample villages was found to be within
the national average ratio (1:7000) in Indonesia, which was again contradicting with the higher
dissatisfaction with long waiting time. It refers to the fact that ratio of health facilities to population
mentioned in policy and the national average ratio of population to doctor were large and unrealistic
when compared to international standards. Although WHO mentions that there are no specific
standards for assessing the sufficiency of medical personnel to address the healthcare need of a given
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
69
population, average ratio of physician to population was estimated (1:1000) in accordance to achieve
adequate coverage rates for primary healthcare. This ratio was met in only four sub districts out of 78
in the DIY. Although the implementation of existing policies regarding availability of one puskesmas
for every 120,000 look optimistic, however if looked into international health standards, the standard
does not seem to be applicable.
Health policy concerning government health insurances was found to be well implemented primarily
focusing on poor population. By providing free primary healthcare service including medicine cost,
issue of equitable access in obtaining adequate primary healthcare was ensured at macro level.
However, service quality in terms of inter personal treatment was a remarkable problematic issue
found in this study. Although poor people are getting free PHC service, they expressed the feeling of
inequality when it comes to personal behaviour of medical staffs in public health facilities. As a
result, majority of people wish to change their health facility to a private clinic if applicable.
If access to PHC was evaluated only considering the factors mentioned in health policy of DIY, then
result might have shown higher level of access than what was found by this study by measuring five
dimensions of access. Hence evaluating access by considering such dimensions have demonstrated
underlying variations that exists in present situation which might not be in policy focus.
6.3 Sub-objective 3: Variations in Results Obtained from Census and Primary
Data
Comparative analyses between aggregated census data and primary household data showed
significant variation in socioeconomic composition of villages. Census presented very small
percentage of households with the letter of poor and health card in sample villages as compared with
the findings from household survey (refer section 5.1). From this observation it can be said that in
comparison, aggregated census data overestimated the socioeconomic condition of sampled
households, over-shading the heterogeneity within them. Such information on census might not reflect
the real situation when it is being used for studies, where socioeconomic characteristics of people play
a vital role. Also such aggregated data might not provide accurate information while making
purposive selection of study area as in this study.
Loss of variation while aggregating can be one of the reasons behind such variation. Also, it could be
that census and primary data were collected in different ways for different purposes. Sample size used
might be small and by chance, may not be a good representative of the villages. Other reasons could
be that the sample villages in this study might not be very representative and could be an extreme
case, although unlikely. Time gap in collection of information could be another factor in such
variation, as the census data was from 2005 and primary data was collected in 2009.
Variation in the geographic distance from villages to nearest health facilities stated in census and that
obtained from survey was expected. As census simply considered distance from village centre to the
nearest facility without consideration to the mode of transportation, assuming people will visit the
nearest one. But in reality, this distance varies as people have their own preference in choosing a
health facility. Decision of not choosing the nearest facility is mostly related to their dissatisfaction
with certain factors in that facility. This could be an indication of some drawbacks in such health
facilities seeking for improvements.
Higher variations within villages than its sub districts refer that intra village variation can get lost in
aggregation process. Hence studying variation in access to healthcare at lower level, than
administratively defined village boundaries can provide better understanding of problems in various
dimensions of access to healthcare for improvement.
70
7. Conclusions and Recommendations
This chapter gives an overview on the main building blocks of this study starting from the
theoretical conceptualisation to the methodological operationalisation in achieving answers to all
research questions formulated in this study. It is composed in two main sections. First section
presents the conclusive remark on the main findings of this study. Second section proposes
recommendations on possible aspects or directions for further research development based on the
findings and limitations of this study.
7.1 Conclusions
This study was primarily concerned with the evaluation of access to primary healthcare considering
various physical, financial and social factors using province of Yogyakarta as a case study. Unlike
common approach of evaluating access by focusing merely on physical accessibility and availability,
this study conceptualised and operationalised access as consisting of five dimensions developed as
follows: accessibility; availability; affordability; acceptability and adequacy. To achieve the main
objective of this study, three supporting objectives were formulated along with number of research
questions. Coming subsection draws conclusion over the findings of this study in order to address the
research questions to achieve the main objective.
7.1.1 Main Findings from Sub-objective 1
- Analyses to address research questions of first sub objective showed that intra village
variation in access to PHC was higher than inter village due to socioeconomic heterogeneity within
each sample villages. Although variation was not large, Tegalpanggung had slightly lower satisfaction
with access when compared to other two villages, as it contained comparatively large proportion of
poor households. Study of higher spatial resolution than at administratively defined area boundaries,
for instance village in this study, can give better insight of the situation on access to PHC. Also if
areas with high socioeconomic homogeneity can be identified for evaluation, then results of analyses
might give clear variation between such areas.
- By quantifying and measuring relevant indicators under each dimension of access, it was
realized that perceived importance of different factors or dimensions in access was related to the
socioeconomic characteristics of individual to a large extent. Healthcare cost seemed more important
than fast service and better personal treatment basically among lower socioeconomic class. As despite
of dissatisfaction with these issues people continued to visit public health facilities for free or cheap
service cost. In contrary, despite of comparably high service and medicine cost, mostly high
socioeconomic class preferred to visit private facilities for faster and better service. Hence
availability and adequacy, as defined in this study, seemed to be more important dimension than
affordability for well off people. Noticeable variation was observed in overall satisfaction level with
access to PHC between different socioeconomic classes. Majority of unsatisfied respondents were
from lower socioeconomic class and satisfied from higher class. However, the satisfaction level was
related to the type of facilities being visited rather that the socioeconomic class. Therefore while
evaluating variation in access to PHC, type of facility should be considered, which can provide better
indication of problematic factors related with certain type of health facility for further improvement.
- Satisfaction level with access to PHC was found to be influenced by satisfaction with
individual factors under dimensions. As in most of cases, despite of higher satisfaction with majority
of factors under five dimensions of access, the overall satisfaction level was lower (34%) than the
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
71
dissatisfaction level (48%) in total respondent. In general, factors related to accessibility,
affordability and acceptability were found to be less problematic in study area, as compared to the
long waiting time under availability and dissatisfaction with inter personal treatment by medical
staffs under adequacy. Only these two factors were found to be associated with the overall satisfaction
level of people visiting public health facilities. Feeling of inequality in personal treatment by medical
staffs under adequacy was expressed by lower socioeconomic class visiting public healthcare, which
was related to their dissatisfaction. Other factors like cost, distance and trust on medical ability or
quality of service showed importance while evaluating access to private health facilities. Short
waiting time in facility, good inter personal treatment and high trust on medical ability seemed to be
important reasons for people to visit private clinics. Hence, their satisfaction level with access had
positive relation with their perception over these issues. This refers to the fact that all these issues are
relevant and should be considered while evaluating access to healthcare in general.
7.1.2 Main Findings from Sub-objective 2
- Second sub objective was to study the health planning system and policy standards in
context of Yogyakarta. Aim was to compare existing situation of access to healthcare at micro level
with the policy standards at macro level. Referring to the health policy of the province of
Yogyakarta, the state of access to PHC was found to be in accordance to the existing policy norms.
Access related policy for healthcare was focused on factors like availability of public health
facilities; physical accessibility; affordability and adequate quality of services. As a result of efficient
policy implementation in these aspects, the existing situation of access was found to be in better state.
Accessibility and affordability ranked higher summary scores, also in perceived level of satisfaction
when evaluated at micro level. Limitation in policy intervention was found in case of perceived
problems with availability of medical personnel and in addressing inequality experienced by people
from different socioeconomic strata in terms of inter-personal treatment by medical staffs.
Availability was perceived as the ratio of population to doctors in this study. Although this ratio was
in line with the average population doctor ratio of Indonesia, it looked very large and inapplicable
when compared with international health standards like WHO. Hence, further attention is required at
policy making level in related problematic issues with consideration of international health standards
if applicable. If issues related to personal treatment along with long waiting time in public health
facilities can be addressed from policy level, then existing access to public healthcare can score
remarkably higher satisfaction level at micro level.
7.1.3 Main Findings from Sub-objective 3
Third sub objective of this study was to evaluate the consistency in results obtained from different
scale of analysis. To address this sub objective, comparative analyses was done between common
socioeconomic and accessibility variables, using aggregated census data from 2005 and
disaggregated primary data obtained from household survey. Large difference in socioeconomic and
accessibility variables was observed. Comparative analyses between socioeconomic attributes
demonstrated that aggregated census data often overestimated the socioeconomic characteristics of
people at village level. Also variation found in socioeconomic and accessibility attributes within
villages were reduced or lost when checked at higher administrative unit than village. These results
indicated that aggregation at larger scale, i.e. village, can average out the variability in
socioeconomic and accessibility variables that exists at small scales, i.e. socioeconomic clusters
within villages. In general, the findings indicate that large scale study might hide the variability in
access to PHC at small scales. If homogeneous socioeconomic clusters can be used as the scale of
analysis rather than census data at village level, then the evaluation of access to healthcare can
72
provide more realistic results. Care should be given to the fact that scale and configuration of spatial
units may affect the outcome of analyses. Therefore care should be taken while drawing conclusion or
in decision making when aggregated data are used.
7.2 Recommendations
Following the findings and methodological limitations faced in this study, number of possible areas
can be recommended for future research development that can further enhance the findings of this
study.
- As this study was carried out in the areas located in city core and immediate periphery of the
city, large variation could not be seen in the state of access to PHC between sample villages. Rural
area with sparsely located settlement can give more space to compare the variation in access between
urban and rural regions within one province under similar health policies. Detail spatial analysis, for
example using road network or gravity based models, can be another practical approach to evaluate
accessibility measures under access.
- Another possible field can be to evaluate variation in access at different scale of analysis
using same source of data. As in this study, comparison in aggregated and disaggregated data was
done using secondary census data and primary household data respectively. Evaluating the
consistency of statistical results obtained from aggregated and disaggregated data will be more
precise and accurate if same source of data can be used.
- Following the result of this study that showed variation in access between different
socioeconomic classes, new zone design technique can be proposed for further research development.
This technique can be used to create new zone boundaries by considering maximum internal
socioeconomic homogeneity. Rather than limiting analyses within administratively defined
boundaries, i.e. village, this technique can show more effective and realistic variations in access to
healthcare across different socioeconomic classes. It also allows to observe the effect of modifiable
areal unit problem.
As a final remark of this study, even the most powerful diagnostic tests, medicines and existence of
supreme quality healthcare service cannot improve peoples‟ health status, if they do not reach to
needy people. To ensure equal access in all dimensions to health services by people, regardless of
their socioeconomic strata, is a major challenge in itself. Hence additional efforts should be made
from policy level to enable all population to gain equitable access to primary healthcare. Also health
policies should be formulated in a way to meet the minimum threshold lines for relevant factors.
Consideration to internationally accepted healthcare standards could be more rational in achieving
adequate access to primary health care.
It is expected that the findings of this study, have been useful to define and measure access to primary
healthcare from broader perspective, highlighting the existing underlying problematic factors for
further improvement and enhancement of access to healthcare in the region.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
73
References
Amer, S. (2007). Towards spatial justice in urban health services planning : A spatial - analytic GIS -
based approach using Dar es Salaam, Tanzania as a case study.299 pp, Utrecht University.
Andersen, R. M., A. McCutcheon, et al. (1983). "Exploring Dimensions of Access to Medical Care."
Bagheri, N., G. Benwell, L., et al. (2005). "Measuring spatial accessibility to primary health care."
Black, M., S. Ebener, et al. (2004). Using GIS to Measure Physical Accessibility to Health Care.
International Health Conference World Health Organization.
Campbell, S. M., M. O. Roland, et al. (2000). "Defining quality of care." Social Science & Medicine
51(11): 1611-1625.
Carstairs, V. (1981). "Small area analysis and health research." Community Medicine 3: 131-139.
Cockings, S. and D. Martin (2005). "Zone design for environment and health studies using pre-
aggregated data." Social Science & Medicine 60(12): 2729-2742.
Das, D. (2008). "Urban Quality of Life: A Case Study of Guwahati." Social Indicators Research
88(2): 297-310.
Dash, L. N. (2000). World Bank and economic development of India. New Delhi, APH Publ. Corp.
ESCAP (2009). Statistical Yearbook for Asia and the Pacific 2008, United Nations.
Field, A. (2005). Discovering statistics using SPSS. London etc., Sage.
Foo, T. S. (2000). "Subjective assessment of urban quality of life in Singapore (1997-1998)." Habitat
International 24(1): 31-49.
Fosu, G. B. (1989). "Access to Health Care in Urban Areas of Developing Societies." Journal of
Health and Social Behavior 30(4): 398-411.
Fosu, G. B. (1989). "Access to Health Care in Urban Areas of Developing Societies."
Frankenberg, E., J. C. Caldwell, et al. (2004). "The effects of access to health care on infant
mortality in Indonesia."
Guagliardo, M. F. (2004). "Spatial accessibility of primary care: concepts, methods and challenges."
International Journal of Health Geographics 3.
Gulliford, M., J. Figueroa-Munoz, et al. (2002). "What does „access to health care‟ mean?" Journal of
Health Services Research and Policy 7(3): 186–188.
Gulzar, L. (1999). "Access to Health Care." Journal of Nursing Scholarship 31(1): 13-19.
Haynes, R., K. Daras, et al. (2007). "Modifiable neighbourhood units, zone design and residents'
perceptions." Health & Place 13(4): 812-825.
Higgs, G. (2005). "A Literature Review of the Use of GIS-Based Measures of Access to Health Care
Services " Health Services and Outcomes Research Methodology 5(2): 119-139.
Hyera, M. B. (2003). Assessment of accessibility to elementary education in informal settlements :
the case of Cebu city, the Philippines. Enschede, ITC: 39.
Ibrahim, M. F. and S. W. Chung (2003). "Quality of Life of Residents Living near Industrial Estates
in Singapore." Social Indicators Research 61(2): 203-225.
74
Lahelma, E., P. Martikainen, et al. (2004). "Pathways between socioeconomic determinants of
health." J Epidemiol Community Health 58: 327–332.
Leisinger, K., M. (2008). Annual report 2008 Creating access – for better health, Novartis
Foundation for Sustainable Development.
Martin, D., A. Nolan, et al. (2001). "The application of zone-design methodology in the 2001 UK
Census." Environment and Planning A 33(11): 1949-1962.
Martinez, J. A. (2005). Monitoring intra - urban inequalities with GIS - based indicators : with a case
study in Rosario, Argentina, Utrecht University ITC.
Maxwell, R. J. (1992). "Dimensions of quality revisited: from thought to action." 1(3): 171-177.
May, R., K. Rex, et al. (2000). "UN Habitat Indicators Database: evaluation as a source of the status
of urban development problems and programs." Cities 17(3): 237-244.
McGrail, M. R. and J. S. Humphreys (2008). "Measuring spatial accessibility to primary care in rural
areas: Improving the effectiveness of the two-step floating catchment area method." Applied
Geography In Press, Corrected Proof.
Millman, M. L. (1993). Access to health care in America, Institute of Medicine.
Moeloek, F. A. (1999). Health Development Plans Towards Healthy Indonesia 2010, Ministry of
Health, Republic of Indonesia.
Moseley, M. J. (1979). Accessibility: the rural challenge, London, Methuen and Co.
Obrist, B., N. Iteba, et al. (2007). "Access to Health Care in Contexts of Livelihood Insecurity: A
Framework for Analysis and Action." PLoS Med 4(10): e308.
Oliver, A. and E. Mossialos (2004). "Equity of access to health care: outlining the foundations for
action."
Omer, I. (2006). "Evaluating accessibility using house-level data: A spatial equity perspective."
Computers, Environment and Urban Systems 30(3): 254-274.
Parker, A. W., J. M. Walsh, et al. (1976). "A Normative Approach to the Definition of Primary
Health Care." The Milbank Memorial Fund Quarterly. Health and Society 54(4): 415-438.
Parnell, S. and M. Poyser (2001). The value of indicators as a tool for local government,
Democratising Local Government: The South African Experiment. University of Cape Town Press,
Cape Town. pp. 251-260.
Penchansky, R. and J. W. Thomas (1981). "The Concept of Access. Definiton and Relationship to
Consumer Satisfaction, Medical Care." 19(2): 127-140.
Rogerson, P. A. (2006). Statistical methods for geography : a student's guide. London etc., Sage.
Roghmann, K. J., A. Hengst, et al. (1979). Satisfaction with Medical Care: Its Measurement and
relation to Utilization, Medical Care.
Schuurman, N., N. Bell, et al. (2007). "Deprivation Indices, Population Health and Geography: An
Evaluation of the Spatial Effectiveness of Indices at Multiple Scales." Journal of Urban Health 84(4):
591-603.
Smith, D. M. (1994). Geography and Social Justice, Colchester, Blackwell.
Stafford, M., O. Duke-Williams, et al. (2008). "Small area inequalities in health: Are we
underestimating them?" Social Science & Medicine 67(6): 891-899.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
75
Talen, E. and L. Anselin (1998). "Assessing spatial equity: an evaluation of measures of accessibility
to public playgrounds." Environment and Planning A 30: 595-613.
Tsou, K. W., Y. T. Hung, et al. (2005). "An accessibility-based integrated measure of relative spatial
equity in urban public facilities." Cities 22(6): 424-435.
Turksever, A. N. and G. Atalik (2001). "Possibilities and Limitations for the Measurement of the
Quality of Life in Urban Areas." Social Indicators Research 53(2): 163-187.
Unger, A. and L. W. Riley (2007). "Slum Health: From Understanding to Action." PLoS Med 4(10):
e295.
Veenhoven, R. (2002). "Why Social Policy Needs Subjective Indicators." Social Indicators Research
58(1): 33-46.
Wagstaff, A. (2002). "Poverty and health sector inequalities." Bulletin of the World Health
Organization 80: 97-105.
WHO (2009). "Indonesia Country Health Profile - Development of the health system." Retrieved 22-
12-2009, 2009, from www.searo.who.int/EN/Section313/Section1520_6822.htm.
Wong, C. (2006). Indicators for urban and regional planning : the interplay of policy and methods :
also e-book. London etc., Routledge.
Yamauchi, F., S. Chowdhury, et al. (2007). "Spatial coordination in public good allocation:
Nonparametric evidence from decentralized Indonesia " International Food Policy Research Institute
(IFPRI)
76
Appendix A: Empirical references to indicators used in this study
Indicators
Pen
chan
sky a
nd
Th
om
as
(1981)
An
der
sen
, M
cCu
tch
eon
et a
l. (
1983)
Fosu
, G
.B.
(19
89)
May,
Rex
et
al.
(2
000)
Wagst
aff
(2002)
Gu
agli
ard
o (
2004)
Fra
nk
enb
erg,
Cald
wel
l
et a
l.
(2004)
Oli
ver
an
d M
oss
ialo
s
(2004)
Un
ger
an
d R
iley
(2007)
Ob
rist
, It
eba e
t al.
(2007)
Am
er (
2007)
Lei
sin
ger
(2008)
Gu
lzar
(1999)
Socio-
economic
Attributes
Household
characteristics
Education level √ √ √ √ √ √ √ √ √ √ √ √
Age dependency √ √ √ √ √
Employment
status /
dependency
√ √ √ √ √
Household
income
√ √ √ √ √ √ √ √ √ √ √ √
Housing
condition
House ownership √ √ √ √ √ √
Land ownership √ √ √
Construction
type
√ √ √ √ √
House crowding √ √ √
Physical
infrastructure
Toilet type √ √ √ √ √ √ √
Source of water √ √ √ √ √ √ √
Electricity √ √ √ √
Sewage disposal √ √ √ √ √
Assets
possession
Type of vehicles √ √ √ √
Telephone √
Television √
Refrigerator √
Dimensions of Access to PHC
Availability Type of PHC
facility
√ √ √ √ √ √ √ √ √ √
PHC facility-
population ratio
√ √ √ √
Doctor -
population ratio
√ √ √ √ √ √
Medical drug
store
√ √ √ √ √
Ease to get
appointment
√ √
Waiting time to
get check-up
√ √ √ √ √ √
Subjective
perception
√ √ √ √ √
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
77
Indicators
Dimensions of Access to PHC
Pen
chan
sky a
nd
Th
om
as
(1981)
An
der
sen
, M
cCu
tch
eon
et a
l. (
1983)
Fosu
, G
.B.
(1989)
May,
Rex
et
al.
(2
000)
Wagst
aff
(2002)
Gu
agli
ard
o (
2004)
Fra
nk
enb
erg,
Cald
wel
l
et a
l.
(2004)
Oli
ver
an
d M
oss
ialo
s
(2004)
Un
ger
an
d R
iley
(2007)
Ob
rist
, It
eba e
t al.
(2007)
Am
er (
2007)
Lei
sin
ger
(2008)
Gu
lzar
(1999)
Accessibility
Travel
distance
√ √ √ √ √ √ √ √ √
Travel time √ √ √ √ √ √ √ √ √ √
Mode of
transport
√ √ √ √ √ √ √ √ √
Subjective
perception
√ √ √ √ √ √
Affordability Possession of
hearth card Or
health
insurance
√ √ √ √ √ √
Insurance
coverage
√ √ √ √ √
Direct and
indirect costs
√ √ √ √ √ √ √ √ √
Subjective
perception
√ √ √ √ √ √ √
Acceptability Religious or
cultural
factors
√ √ √
Adequacy Quality of
medical care
√ √ √ √ √ √ √ √ √
Personal
treatment
√ √ √ √ √ √ √ √
Cleanliness √ √ √ √
Opening hour √ √ √ √
Satisfaction level with access to
health care
√ √ √ √ √
EV
AL
UA
TIO
N O
F A
CC
ES
S T
O P
RIM
AR
Y H
EA
LT
HC
AR
E, C
AS
E S
TU
DY
IN
YO
GY
AK
AR
TA
79
Ap
pe
nd
ix B
: D
es
cri
pti
on
an
d r
ati
on
ale
s o
f in
dic
ato
rs u
se
d i
n t
his
stu
dy
Ind
ica
tors
A
ttri
bu
tes
Des
crip
tio
n
Rati
on
ale
So
cio
-
eco
no
mic
Ho
use
ho
ld
char
acte
rist
ics
Hig
hes
t
educa
tion l
evel
Hig
hes
t le
vel
of
edu
cati
on
att
ain
ed b
y
house
ho
ld m
emb
ers
“Ed
uca
tio
n p
rovi
des
form
al q
ual
ific
atio
ns
that
contr
ibute
to t
he
soci
oec
onom
ic s
tatu
s
thro
ugh
occ
up
atio
n a
nd i
nco
me”
(L
ahel
ma,
Mar
tikai
nen
et
al.
2004).
In t
his
stu
dy i
t
con
trib
ute
s to
ob
serv
e th
e in
fluen
ce o
f ed
uca
tion l
evel
on t
he
atti
tude
of
house
hold
tow
ard
s h
ealt
hca
re a
nd
thei
r per
cepti
on o
n d
iffe
rent
fact
ors
rel
ated
to t
he
acce
ss t
o P
HC
.
Age
dep
enden
cy
Nu
mer
ato
r:
Ho
use
ho
ld m
emb
ers
wit
hin
age
of
16
- 6
0 y
ears
old
Den
om
ina
tor:
To
tal
ho
use
ho
ld m
emb
ers
Peo
ple
bet
wee
n 1
6 –
60
yea
rs (
refe
rrin
g t
o c
ensu
s d
ata
and
off
icia
l re
tire
men
t age
in
Ind
on
esia
) ar
e as
sum
ed t
o b
e ac
tive
and a
ble
to w
ork
for
earn
ing.
House
hold
conta
inin
g
mo
re m
emb
ers
bet
wee
n t
his
age
gro
up c
an h
ave
more
poss
ibil
ity i
n i
mpro
ving t
he
eco
no
mic
sta
tus.
Em
plo
ym
ent
dep
enden
cy
Nu
mer
ato
r:
Fu
ll t
ime
emp
loyee
+ p
art
tim
e em
plo
yee
/ 2
Den
om
ina
tor:
To
tal
ho
use
ho
ld m
emb
ers
Ho
use
ho
ld e
arn
ing i
s der
ived
pri
mar
ily f
rom
pai
d e
mplo
ym
ent.
Ho
use
ho
ld w
ith
lar
ge
num
ber
of
emplo
yed
mem
ber
s ca
n b
e as
sum
ed t
o h
ave
more
ear
nin
g
and
bet
ter
eco
no
mic
sta
tus.
Ran
ge
of
inco
me
Ran
ge
of
ho
use
ho
ld m
on
thly
in
com
e, i
n
Indo
nes
ian
Ru
pia
h
1.
Les
s th
an 7
00
,00
0
2.
70
0,0
00
– 1
,40
0,0
00
3.
1,4
00
,00
0 –
2,8
00
,00
0
4.
2,8
00
,00
0 –
5,6
00
,00
0
5.
Ab
ove
5,6
00
,00
0
Inco
me
det
erm
ines
th
e purc
has
ing p
ow
er o
f house
hold
. It
als
o c
ontr
ibute
s to
res
ourc
es
nee
ded
in
mai
nta
inin
g g
ood h
ealt
h.
Inco
me
bei
ng a
sen
siti
ve m
atte
r, a
ran
ge
is u
sed t
o p
ut
resp
onden
t at
eas
e an
d a
lso t
o a
void
ove
rest
imat
ion
or
un
der
esti
mat
ion t
o s
om
e ex
tent.
(Ref
erri
ng
to
Sta
tist
ics
of
Ind
ones
ia 7
00,0
00 I
nd
ones
ian R
upia
h i
s co
nsi
der
ed a
s th
e
inco
me
po
vert
y li
ne
for
a h
ouse
hold
wit
h 4
-5 m
ember
s.)
Poss
essi
on o
f
„let
ter
of
poor‟
Ask
ing i
f th
e h
ou
seh
old
hav
e th
e le
tter
of
poor
It i
s an
off
icia
l le
tter
iss
ued
by t
he
loca
l gove
rnm
ent
whic
h i
s an
indic
atio
n o
f a
poor
ho
use
ho
ld i
n c
on
tex
t o
f In
dones
ia.
Ho
usi
ng
con
dit
ion
House
and L
and
ow
ner
ship
Sta
tus
of
ho
use
an
d l
and
ow
ner
ship
;
ow
ned
, re
nte
d o
r sq
uat
ted
Ow
ner
ship
of
ho
use
and l
and e
nsu
res
less
fin
anci
al b
urd
en o
n h
ouse
hold
in t
erm
s of
mo
nth
ly r
ent.
Const
ruct
ion
type
Co
nst
ruct
ion
typ
e o
f th
e h
ou
se, if
it
is
per
man
ent,
sem
i p
erm
anen
t o
r te
mp
ora
ry
Ph
ysi
cal
con
dit
ion
an
d t
ype
of
const
ruct
ion c
an e
xpre
ss t
he
livi
ng s
tandar
d o
f house
hold
mem
ber
s.
80
stru
ctu
re
Def
init
ion
of
con
stru
ctio
n t
ypes
fro
m c
ensu
s ques
tionnai
re:
-Per
man
ent
ho
use
is
the
one
wit
h b
rick
, co
ncr
ete
or
wooden
wal
ls,
roof
mad
e out
of
alu
min
ium
sh
eets
or
wooden
sla
tes
and f
loors
wit
h c
oncr
ete
or
cera
mic
til
es.
-Sem
i p
erm
anen
t h
ou
se a
re m
ade
wit
h h
alf
concr
ete,
wood o
r bam
boo w
alls
wit
h r
oofs
mad
e o
f ti
les
or
alu
min
ium
shee
ts o
r w
ood o
r as
bes
tos.
-Tem
po
rary
or
sim
ple
house
are
those
mad
e out
of
mud,
wood a
nd l
eave
s.
C
row
din
g
Rat
io o
f to
tal
nu
mb
er o
f b
edro
om
s to
th
e
tota
l h
ou
seh
old
mem
ber
s
Nu
mb
er o
f p
eop
le p
er b
edro
om
is
calc
ula
ted t
o s
ee t
he
infl
uen
ce o
f bed
shar
ing a
s it
is
con
sid
ered
as
a se
nsi
tive
cro
wdin
g i
ndic
ator
also
for
hea
lth s
tudie
s.
Ph
ysi
cal
infr
astr
uct
ure
Toil
et t
ype
Typ
e o
f to
ilet
use
d, if
it
is p
riva
te, p
ub
lic
(sh
ared
) o
r n
o t
oil
et (
for
exam
ple
, in
rive
r)
Bas
ic i
nfr
astr
uct
ure
s li
ke
wat
er s
upply
, sa
nit
atio
n a
nd e
lect
rici
ty a
re c
om
monly
use
d
ind
icat
or
for
soci
oec
onom
ic s
trat
ific
atio
n.
Bei
ng b
asic
nee
ds
for
dai
ly a
ctiv
itie
s, t
hes
e in
fras
truct
ure
s re
flec
t th
e li
ving s
tandar
d a
lso
the
hygie
ne
con
dit
ion
of
house
hold
mem
ber
s. A
s so
urc
e of
dri
nkin
g w
ater
supply
and t
ype
of
san
itat
ion
is
dir
ectl
y r
elat
ed w
ith h
ealt
h s
tatu
s.
Sourc
e of
wat
er
If h
ou
seh
old
hav
e p
ub
lic
pip
ed w
ater
supp
ly (
PA
M),
wel
l, ra
inw
ater
or
fro
m
rive
r as
th
e so
urc
e o
f w
ater
Ele
ctri
city
If
ho
use
ho
ld h
ave
pu
bli
c el
ectr
ic s
up
ply
(PL
N),
so
me
oth
er t
yp
e o
r n
o e
lect
rici
ty
Ass
ets
po
sses
sio
n
Num
ber
of
vehic
le
Num
ber
of
bic
ycl
es, m
oto
r b
ikes
or
cars
A
sset
po
sses
sio
n i
s an
oth
er s
oci
oec
onom
ic i
ndic
ator
whic
h i
ndic
ates
the
stat
us
and a
bil
ity
of
ho
use
ho
ld t
o m
ain
tain
cer
tain
lev
el o
f li
ving s
tandar
d.
As
inco
me
bei
ng a
sen
siti
ve a
nd s
om
etim
es c
onfi
den
tial
mat
ter,
eva
luat
ion o
f su
ch a
sset
s
can
be
a u
sefu
l w
ay f
or
soci
oec
onom
ic s
trat
ific
atio
n.
Num
ber
of
phone;
TV
;
refr
iger
ator
Num
ber
of
such
ho
use
ho
ld a
sset
s o
wn
ed
by t
he
ho
use
ho
ld
Acc
ess
to P
HC
Ava
ilab
ilit
y
Deg
ree
of
fit
bet
wee
n
exis
tin
g P
HC
serv
ices
and
nee
ds
of
peo
ple
Type
of
PH
C
faci
lity
Typ
e o
f P
HC
fac
ilit
y v
isit
ed, if
it
is
pu
bli
c(go
vern
men
t) o
r p
riva
te
org
aniz
atio
n
Typ
e o
f P
HC
fac
ilit
y b
eing v
isit
ed c
an b
e use
ful
to e
valu
ate
resp
onden
t‟s
per
cepti
on
tow
ard
s va
rio
us
fact
ors
rel
ated
to h
ealt
hca
re s
ervi
ce.
It c
an a
lso e
xpre
ss t
he
soci
oec
onom
ic
stat
us
of
resp
on
den
t. A
s in
conte
xt
of
Indones
ia,
gove
rnm
ent
hea
lth c
entr
es p
rovi
de
free
PH
C s
ervi
ce t
o p
oo
r p
eople
.
PH
C f
acil
ity-
popula
tion r
atio
Num
ber
of
PH
C f
acil
itie
s d
ivid
ed b
y
tota
l p
op
ula
tio
n i
n e
ach
su
b-d
istr
ict
Th
e h
ealt
h p
oli
cy i
n I
nd
ones
ia h
ave
spec
ifie
d a
min
imum
num
ber
of
hea
lthca
re f
acil
itie
s
per
ad
min
istr
ativ
e u
nit
s
Th
e ra
tio
of
faci
lity
or
doct
ors
to p
opula
tion s
how
s th
e dem
and
-supply
rat
io o
f P
HC
serv
ice.
It
also
en
able
s to
com
par
e th
e ex
isti
ng s
ituat
ion w
ith t
he
DIY
poli
cy a
nd
Doct
or
-
po
pula
tion r
atio
Tota
l n
um
ber
of
do
cto
rs d
ivid
ed b
y t
ota
l
popu
lati
on
in
eac
h s
ub
-dis
tric
t
EV
AL
UA
TIO
N O
F A
CC
ES
S T
O P
RIM
AR
Y H
EA
LT
HC
AR
E, C
AS
E S
TU
DY
IN
YO
GY
AK
AR
TA
81
inte
rnat
ion
al h
ealt
h s
tandar
ds.
Med
ical
dru
g
store
Ava
ilab
ilit
y o
f m
edic
al s
tore
in
PH
C
faci
lity
vis
ited
Ava
ilab
ilit
y o
f d
rug s
tore
for
pre
scri
bed
med
icin
e is
an i
mport
ant
fact
or
consi
der
ed i
n
man
y e
mp
iric
al s
tud
ies
(Wag
staf
f 2002;
Am
er 2
007;
Obri
st,
Iteb
a et
al.
2007)
whil
e
eval
uat
ing a
cces
s to
hea
lthca
re.
It i
s as
sum
ed t
hat
peo
ple
s‟ o
pin
ion o
n o
vera
ll a
vail
abil
ity
of
hea
lth
care
is
infl
uen
ced b
y h
ow
eas
y o
r dif
ficu
lt i
t is
to g
et r
equir
ed m
edic
ine
is.
Eas
e to
get
appoin
tmen
t
Do t
hey
hav
e to
get
an
ap
po
intm
ent
pri
or
visi
tin
g t
he
faci
lity
or
just
wal
k i
n?
If
yes
, h
ow
lo
ng d
oes
it
tak
e to
get
it?
Qu
ery a
bo
ut
app
oin
tmen
t sy
stem
aid
to u
nder
stan
d t
he
pro
cess
by w
hic
h p
eople
can
ente
r
dif
fere
nt
hea
lth
care
fac
ilit
y t
o o
bta
in t
he
serv
ice.
Als
o t
o k
now
how
long t
hey
hav
e to
wai
t
bef
ore
act
ual
ly e
nte
rin
g t
he
faci
lity
, if
appoin
tmen
t is
nee
ded
.
Wai
ting t
ime
to
get
chec
k-u
p
Wai
tin
g t
ime
afte
r re
gis
trat
ion
in
th
e
faci
lity
til
l th
e v
isit
to
do
cto
r fo
r ch
eck
-
up
Wai
tin
g t
ime
in h
ealt
hca
re f
acil
ity i
s an
im
port
ant
and c
om
monly
use
d i
ndic
ator
in
hea
lth
care
rel
ated
stu
die
s. D
espit
e of
physi
cal
avai
labil
ity o
f se
rvic
e ce
ntr
e, p
atie
nts
mig
ht
hav
e to
wai
t lo
ng t
o o
bta
in m
edic
al s
ervi
ce d
ue
to i
nad
equat
e per
sonnel
, eq
uip
men
t or
som
e o
ther
rea
son
. H
ence
, th
is i
ndic
ator
can g
ive
an i
ndic
atio
n o
f re
lati
on b
etw
een v
olu
me
of
hea
lth
care
ser
vice
dem
and a
nd s
upply
.
Su
bje
ctiv
e p
erce
pti
on
on w
aiti
ng t
ime
is r
elev
ant
in t
his
stu
dy a
s it
was
ass
um
ed t
hat
per
cep
tio
n m
igh
t d
iffe
r dep
endin
g o
n s
oci
oec
onom
ic c
har
acte
rist
ics
of
peo
ple
and a
lso
dep
end
ing o
n t
he
typ
e o
f hea
lthca
re f
acil
ity v
isit
ed.
Subje
ctiv
e
per
cepti
on
Opin
ion
ab
ou
t th
e w
aiti
ng t
ime
in
faci
lity
bef
ore
get
tin
g c
hec
ku
p i
n f
ive
Lik
ert
scal
e; V
ery s
ho
rt. S
ho
rt, N
orm
al,
Lon
g a
nd
Ver
y l
on
g
Acc
essi
bil
ity
Deg
ree
of
fit
bet
wee
n
geo
gra
phic
al
loca
tio
n
of
PH
C s
ervi
ce
an
d th
e
loca
tio
n o
f
peo
ple
Is t
he
visi
ted
faci
lity
the
nea
rest
one
Do t
he
ho
use
ho
ld v
isit
th
e n
eare
st P
HC
faci
lity
It w
as a
ssu
med
th
at p
eople
are
rat
ional
and v
isit
s th
e nea
rest
PH
C f
acil
ity i
n i
dea
l ca
se.
Th
is i
nd
icat
or
hel
ps
to e
nsu
re t
he
assu
mpti
on.
Tra
vel
dis
tance
P
hysi
cal
dis
tan
ce t
o t
he
PH
C f
acil
ity
visi
ted
Tra
vel
dis
tan
ce a
nd
tim
e ar
e th
e co
mm
only
use
d i
ndic
ators
to e
valu
ate
trav
el i
mped
ance
in
any a
cces
sib
ilit
y r
elat
ed s
tudie
s. I
t ca
n b
e use
d t
o a
nal
yze
how
far
or
how
long p
eople
are
wil
lin
g t
o t
rave
l to
get
PH
C s
ervi
ce.
Mo
de
of
tran
spo
rt i
s an
oth
er f
acto
r to
be
consi
der
ed i
n t
his
stu
dy a
s it
is
dir
ectl
y r
elat
ed t
o
ph
ysi
cal
acce
ssib
ilit
y. F
or
inst
ance
, non
-moto
rize
, m
oto
rize
d o
r publi
c tr
ansp
ort
atio
n.
Tra
vel
tim
e T
ota
l tr
avel
lin
g t
ime
to r
each
th
e P
HC
faci
lity
Mode
of
tran
sport
Mod
e o
f tr
ansp
ort
use
d t
o v
isit
PH
C
faci
lity
Subje
ctiv
e
per
cepti
on
Opin
ion
ab
ou
t th
e tr
avel
dis
tan
ce a
nd
trav
el t
ime
to r
each
th
e P
HC
fac
ilit
y
Su
bje
ctiv
e p
erce
pti
on
on t
rave
l im
ped
ance
can
hel
p i
n u
nder
stan
din
g b
ehav
ioura
l
dif
fere
nce
in
var
iou
s d
emogra
phic
or
soci
oec
onom
ic c
lass
es o
f peo
ple
.
Aff
ord
abil
ity
Deg
ree
of
fit
Poss
essi
on o
f
hea
rth c
ard
If t
he
ho
use
ho
ld h
ave
the
hea
lth
car
d,
„Kat
u S
ehat
‟
„Kat
u S
ehat
‟ is
a t
yp
e of
gove
rnm
ent
hea
lth i
nsu
rance
pro
vidin
g f
ree
PH
C s
ervi
ce w
hic
h
has
a d
irec
t re
lati
on
wit
h t
he
dim
ensi
on a
fford
abil
ity i
n t
his
stu
dy.
Insu
rance
If
th
e h
ealt
h c
ard
co
vers
med
ical
an
d
Ap
art
fro
m d
irec
t se
rvic
e co
st l
ike
doct
or‟
s fe
e, t
her
e ar
e m
any i
ndir
ect
expen
ses
in
82 bet
wee
n P
HC
serv
ice
cost
an
d
peo
ple
s’
abil
ity
or
wil
lin
gn
ess
to
pa
y
cove
rage
lab
ora
tory
ex
pen
ses
hea
lth
care
wh
ich
sh
ou
ld n
ot
be
ignore
d w
hil
e ev
aluat
ing o
ver
all
acce
ss t
o P
HC
.
Un
der
stan
din
g o
f in
sura
nce
cove
rage
(hea
lth c
ard i
n t
his
cas
e) o
f th
ese
indir
ect
expen
ses
is
rele
van
t, a
s it
can
co
mp
lete
ly i
nfl
uen
ce p
eople
s‟ o
pin
ion o
n h
ealt
hca
re c
ost
.
Dir
ect
and
indir
ect
cost
s
Co
st o
f re
gis
trat
ion
, d
oct
or‟
s fe
e,
med
icin
e, l
abo
rato
ry a
nd
tra
vel.
Subje
ctiv
e
per
cepti
on
Opin
ion
ab
ou
t th
e o
vera
ll c
ost
fo
r P
HC
Acc
epta
bil
ity
Deg
ree
of
fit
bet
wee
n t
he
cha
ract
eris
tics
of
the
pro
vid
er
an
d t
ho
se o
f
peo
ple
Rel
igio
us
or
cult
ura
l fa
ctors
Do r
esp
on
den
t h
ave
any c
ult
ura
l o
r
reli
gio
us
pre
fere
nce
to
vis
it c
erta
in P
HC
faci
lity
?
Peo
ple
mig
ht
hav
e ce
rtai
n r
elig
ious
or
cult
ura
l pre
fere
nce
in v
isit
ing a
par
ticu
lar
hea
lthca
re
faci
lity
, ei
ther
bec
ause
of
som
e so
cial
beh
avio
ur
or
per
sonal
wis
h.
For
inst
ance
, peo
ple
no
rmal
ly w
ish
to
vis
it a
fac
ilit
y w
her
e th
ey c
an c
om
munic
ate
in l
oca
l la
nguag
e. S
imil
arly
,
gen
der
pre
fere
nce
is
anoth
er f
acto
r in
acc
epta
bil
ity,
as i
n s
om
e ca
se f
emal
e pat
ients
wis
h
to b
e ch
eck
ed b
y f
emal
e doct
ors
and m
ale
pat
ients
by m
ale
doct
ors
.
Gen
der
pre
fere
nce
Is t
her
e an
y g
end
er p
refe
ren
ce o
f
doct
ors
?
Ad
equ
acy
Deg
ree
of
fit
bet
wee
n
peo
ple
s’
exp
ecta
tion
an
d s
ervi
ce
pro
vid
ed
Tru
st o
n m
edic
al
qual
ity
Opin
ion
ab
ou
t th
e m
edic
al a
bil
ity a
nd
serv
ice
qu
alit
y o
f th
e vi
site
d P
HC
faci
lity
?
Des
pit
e o
f sh
ort
ph
ysi
cal
dis
tance
to h
ealt
hca
re f
acil
ity o
r bei
ng p
rovi
ded
wit
h f
ree
hea
lth
care
ser
vice
, p
eople
mig
ht
not
be
hap
py w
ith t
he
serv
ice
due
to s
ever
al o
ther
rea
sons.
Fo
r ex
amp
le p
oo
r se
rvic
e qual
ity c
om
par
ed t
o o
ther
hea
lthca
re f
acil
itie
s, u
nfr
iendly
per
son
al t
reat
men
t b
y m
edic
al p
erso
nnel
, unple
asan
t or
dir
ty p
hysi
cal
envi
ronm
ent
etc.
Op
enin
g h
ou
r o
f h
ealt
hca
re s
ervi
ce s
hould
als
o b
e co
nsi
der
ed a
s peo
ple
mig
ht
hav
e
pro
ble
m v
isit
ing t
he
faci
lity
due
to o
verl
appin
g i
n o
pen
ing t
ime
of
the
faci
lity
and t
hei
r
wo
rkin
g (
emp
loym
ent)
sch
edule
.
Su
ch i
nd
icat
ors
are
rel
evan
t in
this
stu
dy a
s th
ey i
nfl
uen
ce t
he
ove
rall
sat
isfa
ctio
n l
evel
of
peo
ple
to
war
ds
acce
ss t
o P
HC
.
Per
sonal
trea
tmen
t
Opin
ion
ab
ou
t th
e p
erso
nal
tre
atm
ent
by
faci
lity
per
son
nel
in
5 L
iker
t sc
ale
(Ver
y
good
, go
od
, n
orm
al, b
ad a
nd
ver
y b
ad)
Cle
anli
nes
s O
pin
ion
ab
ou
t th
e cl
ean
lin
ess
or
physi
cal
app
eara
nce
of
the
visi
ted
fac
ilit
y
Open
ing h
our
Open
ing h
ou
r o
f th
e P
HC
fac
ilit
y a
nd
if
it s
uit
s th
eir
wo
rkin
g h
ou
r
Su
bje
ctiv
e
per
cep
tio
n o
n
ove
rall
acc
ess
to P
HC
Sat
isfa
ctio
n l
evel
T
akin
g a
ll t
he
fact
ors
an
d d
imen
sio
ns
of
acce
ss i
nto
co
nsi
der
atio
n, o
pin
ion
on
th
e
ove
rall
sat
isfa
ctio
n l
evel
to
war
ds
acce
ss
to P
HC
fac
ilit
y
Rat
ion
al o
pin
ion
on
sat
isfa
ctio
n l
evel
tow
ards
acce
ss t
o P
HC
can
ref
lect
the
actu
al
situ
atio
n, in
ter
ms
of
peo
ple
s‟ p
erce
pti
on,
of
acce
ss t
o P
HC
tak
ing e
ach d
imen
sion i
n
acco
un
t.
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
83
Appendix C: Content of household survey questionnaire
Sub-district: ………………………………..….
Interviewer name: …………………………….
Village: …………………………………………
Date: ……………………………………..…..…
House no. : ……………………………..…….. Duration: ……………………….…………..….
This survey is purely for study purposes to understand the existing condition in accessing health care facilities in
the city of Yogyakarta, Indonesia. The individual perception of citizens is of immense value for a successful
completion of this study. Your responses to this questionnaire will be treated with strict confidentiality. Hence,
your honest comments and cooperation will be highly appreciated.
Socio-economic Information
[* Read: I will start this interview with some questions related to your household information.]
A. General information of respondent
A.1.
Respondent is
Male Female
A.2. Position in family Head of family Yes No
Husband Wife
A.3. Household members
[Tick the box for those
who are currently living in
the house from atleast
past 1 year]
Below age 6 years: ………………………..
Age 7 – 15 years: …….……………………
Age 16 – 45 years: ………………………..
Age 46 – 60 years: ………………………..
Above 60 years: …………………………..
Total number (including respondent): ………..…
A.4 Highest education level in
household
No education Elementary school
Junior high school Senior high school
University education Others
A.5. Employment status of
household members
Household member
(eg. husband / wife / son)
Full time job Part time job
…………………….
…………………….
…………………….
A.6. Average monthly income
of household in
Indonesian rupiah
Less than 700,000
700,000 – 1,400,000
1,400,000 – 2,800,000
2,800,000 – 5,600,000
Above 5,600,000
A.7. Do you have „letter of
poor‟ (Surat Keterangan
Tidak Mampu)?
Yes No
A.8. In which socio-economic
group your household
belongs to?
Sangat kaya (High class) Kaya (Upper middle class)
Menengah (Middle class) Miskin (Lower middle class)
Sangat miskin (Poor)
84
[* Read: Now, I will ask some general questions about your house condition.]
B. Housing Condition
B.9. Status of house Owned Rented Other
B.10
.
Status of land Owned Rented Squatted (informal)
B.11
.
Number of rooms
occupied by household
Bedroom: …………………....… Kitchen: ……………..……….
Toilet / Bathroom: …………..… Others: ………………….……
B.12
.
Construction type Type: Permanent Semi-permanent Temporary
B.13
.
Type of toilet Public Private Open air (no toilet)
Shared Not shared
B.14
.
Wastewater disposal
(sewage waste)
Septic tank Sewer line None Other
B.15
.
Trash disposal (garbage) Garbage collection system (……..……………….how often)
In container Burn Roadside
In river Other ……………….…………..……..
B.16
.
Main source of water
supply in house
PAM piped water to house Buy from vendors
PAM piped water in public tap Ground water (well)
Others………………………………………………………
B.17
.
Which source of
electricity supply you
have in house?
PLN (State electricity company) Local government agency
Private corporation Public self reliance agency
No electricity Others………………………
B.18
.
Which of these assets
does your household
have?
Cycle………………….. Telephone……….………….
Motorbike……………… Television…………....….….
Car…………………….. Refrigerator…………....……
Other mode of transport……………………..
Existing health care facility
[* Read: I will now ask you about health condition in your household and health facility you visit.]
C. General information on access to health care
C.19.
In the last 6 months,
how many times have
your household
members suffered
from diseases
mentioned in the
table at right box?
Disease
Members Age How
many
times
Severe flu (cough, cold)
Fever
Diarrhea & vomiting
Malaria
Measles (Campak)
Dengue fever (Demam Berdarah)
Upper Respirator Tract Infection (ISPA)
Other……………………………………..
C.20. ** Which health
facility do your
household members
visit for primary
health care?
(Tick the boxes) Order of visit
Hospital …………
Health care (Puskesmas) …………
Health sub-care (Puskesmas Pembantu) …………
Integrated health posts (Posyandu) …………
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
85
If more than one then
please state in order,
which one do you
visit first and then
further, starting from
1 onwards.
Midwife (Bidan) …………
Traditional birth attendance (Paraji) …………
Others …………………………………………
What is the type of facility?
Public (government) Private
Type of treatment.
By professional medical personnel (like doctor)
By traditional practitioner without formal medical education
Name of facility:………… ……………………………………………
C.21. Is the facility visited
by your household,
the nearest one?
Yes No
If no, why didn‟t you go to the nearest one?
More expensive Don‟t trust the ability of doctors
Unfriendly behavior Don‟t like the quality of service
Religious factor Other………………….……………
C.22. How far is the
facility that your
household visits?
Distance: ………………………….
Time to reach: ……….…………… (in normal condition)
…………………… (with traffic jams, if any)
C.23. Which mode of
transport is used to
reach the facility?
Walk Bicycle Motorbike Private car
Public transportation
If public transport, is it easily available? Yes No
How long do you have to wait to get it? ............................... (time)
C.24. What do you think
about the distance to
the facility?
Very near Near Normal Far Very far
C.25. What do you think
about the travel time
to reach the facility?
Very short Short Normal Long Very long
C.26. Is it necessary to get
an appointment?
Yes No
If yes, can it be done by telephone call?
Yes No
Getting appointment is:
Very easy Easy Normal Difficult Very difficult
C.27. How long do you
normally have to wait
to get a checkup,
after reaching the
health facility?
Time: ……………………………………………..
What do you think about the waiting time?
Very short Short Normal Long Very long
Does it have proper waiting area? Yes No
C.28. Does the facility
have a medical shop
providing prescribed
medicines?
Yes No
If no, how far do you have to go to buy the prescribed medicine?
Very near Near Normal Far Very far
C.29. Does your household
have a health card
(Kartu Sehat) or
health insurance?
Yes No
If yes, who provides it? Government Private organization
Other……………………………………
If no, Why?
86
Not eligible Long process to get the card
Have to travel long distance Unaware
Don‟t need it Other………………………..
C.30. How much do you
have to pay to get
health insurance?
………………………………...……Rupiah Free
If not free, what do you think about the cost?
Very inexpensive Inexpensive Normal
Expensive Very expensive
C.31. ** Does the
insurance cover all
health care
expenses?
Yes No Only primary health care
If no, up to what amount does it cover for bigger diseases?
……………………………..........................
Does it cover all medicine expense? Yes No
C.32. In which health
facilities does the
health card or
insurance give free
health check up?
(See C.20. for
reference)
Government (Public) Private
Hospital
Puskesmas
Puskesmas Pembantu
Posyandu
Midwife (Bidan)
Traditional birth attendance
Hospital
Puskesmas
Puskesmas Pembantu
Posyandu
Midwife (Bidan)
Traditional birth attendance
All
Others ………………………………………………….…………
C.33. How much additional
cost does your
household spend in
health care?
………………… Registration cost ………………… Doctor‟s fee
………………… In medicines …….……….…… Travel cost
…………….…… Overall
C.34.
What does your household think about these costs?
Doctor‟s fee Very inexpensive Inexpensive Normal
Expensive Very expensive
Medication cost Very inexpensive Inexpensive Normal
Expensive Very expensive
Travel cost to get
health care
Very inexpensive Inexpensive Normal
Expensive Very expensive
Total cost Very inexpensive Inexpensive Normal
Expensive Very expensive
C.35. Does your household
feel welcome in the
facility you visit?
Yes No
If no, how and why
…………………………………………………………………..………
C.36. Is there any cultural or
religious preference in
choosing a particular
facility?
Yes No
If yes, what……………………………………………………………
Very satisfied Satisfied Normal
Unsatisfied Very unsatisfied
C.37. How is the cleanliness
of the facility?
Very Clean Clean Normal Dirty Very dirty
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
87
C.38.
How is the personal
treatment from all
facility personnel?
Very Good Good Normal Bad Very bad
C.39.
What does your
household think about
medical ability (trust)
of the facility?
Very Good Good Normal Bad Very bad
**C.40
.
If an equal number of
male and female
medical personnel is
available, to whom
will your household
prefer to visit for
check up?
Male household members to male doctor
Female household members to female doctor
Does not matter for male female
Does not matter at all
How satisfied are you with existing situation in this regard?
Very satisfied Satisfied Normal
Unsatisfied Very unsatisfied
C.41. Does the opening
hour of the facility
suits your household
time?
Yes No
C.42.
If your household
income is doubled,
will your household
still go to the same
health care facility?
Yes No
If no, then which one (name)
………………………………………………………………….…….…
Why?………………………………………………………….………...
**C.43
.
Which of these factors
is more important for
you to get better
health care?
Please rank your preference from 1 to 6
Waktu tempuh (Reduced travel time) ………
Waktu tunggu (Reduced waiting time) ………
Biaya (Reduced cost) ………
Faktor budaya dan agama (Cultural / religious factors) ………
Kualitas yang baik (Improved quality of service) ………
Keramahan personel (Friendliness of facility personnel) ………
C.44. Over all satisfaction
level towards existing
health care that you
are getting.
Very satisfied Satisfied Normal
Unsatisfied Very unsatisfied
C.45. How do you think
access to health care
can be improved?
Mengurangi waktu perjalanan (Further reduced travel time)
Mengurangi waktu tunggu (Further reducing waiting time)
Mengurangi biaya (Further reduced cost)
Banyak pilihan dengan mempertimbangkan factor agama
dan tradisi (Better options regarding cultural factors)
Meningkatkan kualitas jasa pelayanan (Further improving quality
of service)
Meningkatkan personal dari pelayan rumah sakit keramahan
(Further improving personal treatment by facility personnel)
Thank you very much for your time and cooperation!!
88
Appendix D: Content of Interview with Health Facility Personnel
Sub-district: ………………….……………….
Interviewer name: ………………………..……….
Village: …………………………….………….
Position of interviewee: …………………......……
…………………………………………….……
Name of facility: …………………….………..
Type of facility:………..……………………………
Location : …..……………………………….…
(Coordinates:………………………………..…….)
Date: ……………………….…
Duration: ……………….…….
This survey is purely for study purpose to understand the existing condition in accessing health care facility in
the city of Yogyakarta, Indonesia. Individual perception of citizen is of immense value for a successful
completion of this study. Your responses to this questionnaire will be treated with strict confidentiality. Hence,
your honest comments and cooperation will be highly appreciated.
[* Read: I will ask some general questions related to the service of this health care facility]
A. General information from facility personnel
A.1. In which year did this
service start?
………………………………………………………………….
A.2. Type of service Primary examination Curative care Maternity care
Dental care Inpatient All Other
A.3. Number of medical
doctors
………………………Male
………………………Female
A.4 Number of nurse or other
medical assistants
………………………Male
………………………Female
A.5. Number of patients‟ bed ……………………… for general primary care
……………………… for emergency case
……………………… for indoor patients (admitted for few days)
A.6. Does this facility have all
laboratories and
equipments required for
primary health care?
Yes No
A.7. Does this facility provide
any extra service or
treatment equipment that
others don‟t have?
Yes No
If yes, what…………………………………………………...……..
………………………………………………………………………..
A.8. Is appointment required?
Yes No
By telephone? Yes No
If yes, before………………………………….. hours or days.
A.9. Number of patient
attendance in last month
………………………………………………… (total)
………………………………………………… ( from poor family)
EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA
89
A.10. Number of patient
attendance in last year
………………………………………………… (total)
………………………………………………… (from poor family)
A.11. Attendance of poor
patients in last 6 months
suffering from mentioned
diseases
Disease
Total
number
Number of poor patients
with card
Severe flu (cough, cold)
Fever
Diarrhea & vomiting
Malaria
Measles
Dengue fever
Upper Respiratory Tract
infection (ISPA)
Others
…………………………
A.12. Opening hour ……………………………………………………..
A.13. Does this facility provide
emergency service?
Yes No
Number of ambulance…………………… Don‟t have it
A.14. Where do majority of
patients come from?
…………………..….... Village ..………….………..Sub-district
…………………………. Regency
A.15. Cost Registration cost:……………………………………………….….
Doctor‟s fee (for primary care)n .....……………………….………
Other (if any): ……………………………………………………...
A.16. Do patients have problem
in paying?
Yes
No
A.17. Any subsidized rate of
treatment for special group
of people?
Yes
No
If yes, who……………………………………………………………
(with health card , letter of poor or others)
A.18. How far does the health
insurance cover the cost?
Cover all the cost
Cover partially (……………..……..% of total health care cost)
Does health insurance cover medication cost?
Yes No
A.19. Do patients have any
preference for male or
female doctors / medical
personnel?
Yes
No
A.20. Do patients have any
common complain?
Yes
No
If yes, what……………………………………………………….
Self observation
Amount of patients in waiting lobby…………………………………………………....…
Enough seating provision in waiting lobby…………………………………………...….…
Cleanliness ……………………………………………………………………………....……
Availability of medical store…………………………………………………………......……
General mode of transport used by patients……………………………………………..…….
Distance to nearest public transport transit …………………………………………...……
90
Appendix E: Distribution of summary scores of dimensions of
access
Figure E-1displays the distribution of summary scores for each dimension using box plot which depict
the median, quartile and outliers. The outliers are the household with values between 1.5 inter
quartile range (IQR) and 3 IQR‟s from the end value of a box. Outliers with low accessibility scores
in box plot showed in Figure (a) represented households visiting hospitals from village Tridadi.
And the outliers with higher adequacy scores were households from higher socioeconomic class who
visited private clinic from village Kricak. Figure (b) shows the distribution of summary scores for
households visiting three different health facilities. Distribution of accessibility score for hospital and
affordability score for private clinics showed wide range with low median value as compared to other
two. Adequacy score was much higher for private clinics.
Table E-1 presents the summary scores of dimensions of access for sample villages and different
healthcare facilities.
Accessibility Availability Affordability Acceptability Adequacy
Village
Tegalpanggung
Kricak
Tridadi
0.92
0.87
0.78
0.73
0.74
0.7
0.78
0.83
0.8
0.84
0.79
0.79
0.72
0.7
0.75
Healthcare facility
Hospital
Puskesmas
Sub-puskesmas
Private clinic
0.71
0.79
0.8
0.76
0.76
0.73
0.74
0.82
0.75
0.8
0.82
0.67
0.82
0.76
0.77
0.81
0.74
0.65
0.66
0.79
Table E-1: Summary scores of dimensions of access for sample villages and different
healthcare facilities
[Note: High values indicate high level of access; 1 = maximum value]
Figure E-1: Distribution of dimensions summary scores [High score refers to higher satisfaction level; 1= highest score]
(b) Score per health facilities
1.0
0.8
0.6
0.4
0.2
0 Hospital Puskesmas Private clinic
Accessibility Availability Affordability
Acceptability Adequacy
(a) Total sample score
1.0
0.8
0.6
0.4
0.2
0
Accessibility Availability Affordability Acceptability Adequacy